ANTAKI - Fitness For Service and Integrity of Piping-Vessels and Tanks PDF

A COMPLETE GUIDE TO ASME RECOMMENDED PRACTICES
FOR ASSESSING, MONITORING, ANO MAINTAINING

PRESSURE VESSELs-AND PIPING
Fitness-for-Service and lntegrity of Piping, Vessels, and Tanks provides

o ~
,...... .
~
instruction on the latest ASME recommended practices for inspecting, evalu- t"""''"-
ating, and monitoring pressure vessels and piping.
This handy volume annotates and explains ASME code and API inspection
and fitness-for-service practices. Presented in a clear, precise manner, this
~ ~
,......
~
,......
.
('D
.
U'J.
Fitness-for- Service
U'J.
and lntegrity of
comprehensive guide meticulously explains how stress analysis, materiais,
engineering, and nondestructive inspection affect fitness-for-service.
•
The tools needed to ...
lnspect and categorize damage, degradation, and corrosion
CJq
~
...., oJ
~
I
•
mechanisms
Evaluate wall thinning, pitting, cracking, fatigue, mechanical
damage, high-temperature corrosion, and creep
~
U'J.
U'J.
I
[fJ
('D
~
Piping, Vessels,
•
•
Develop and implement programs for corrective and predictive
maintenance of tanks, vessels, piping, and pipelines
Select the right rep air technique and avoid pitfalls in repair weld ing
and joining
('D
~
__,U'J.
<.
,......
n
('D
and Tanks
~
~ ~
~
O- O-
A COMPREHENSIVE GUIDE TO ASME RECOMMENDED PRACTICES: ASME CODE SIMPLIFIED
• What Causes Failures • Material Defects • Design Errors
~ ~
~
~
• Construction Flaws • Corrosion and Damage Mechanisms
t"""''"-
• Fai/ures ln-Service • lnspection Techniques
~('D
• Maintenance and Risk-Based lnspection Strategies oo(}q
• Failure Margins and Prevention • Margins to Failure
• Run-or-Repair Decisions • Repair Techniques
~
,...... .
~ V Run-or-repair evaluations
The McGrow·Hi/1 Componles -., _ ASME code for operating equipment and systems
Antaki V ASME and API rules for structural integrity
Visit McGraw-Hill Foucation at

www.books.mcgraw·hill.eom
Mechanical Engineering
George Antaki
,. -
Other Books of lnterest from McGraw-Hill ~
+
ELLENBERGER • Piping Systems and Pipeline: ASME Code Simplified t Fitness-for-Service
ELLENBERGER • Pressure Vessels: ASME Code Simpli{ied, Eighth Edition
MENON • Piping Calculations Manual
PARMELY • Machzne Devices and Components
l
•
T
I and lntegrity of
WOODRUFF, LAMMERS, LAMMERS • Steam Plant Operations, Eighth Edition
Piping, Vessels,
li andTanks
ASME Code Simplified
i
1 George Antaki
f
t
.!r
McGraw-Hill
New Yor k Chicago San Francisco lisbon London Madrid
Mexlco City Milan New Oelhi San Juan Seoul
Singapore Sydney Toronto
The McGraw·Hill Companles ,·
Contents
Cataloging-in-Publication Data is on file with the Library of Congress.
Copyright © 2005 by The McGraw-Hill Companies, Inc. All rights reserved.

Printed in the United States of America. Except as permitted under the United
States Copyright Act of 1976, no part of this publication may be reproduced or Preface xlll
distributed in any fonn or by any means, or stoxed in a data base or retrieval
system, without the prior written permission of the publisher.
Chapter 1. Principies 1
1234567890 oocmoc 01098765 1.1 What ls Fitness-for-Service?
1.2 FFS and Conduct of Operations 1
ISBN 0-07-145399-7 1.3 Fitness-for-Service of Old and New Equipment 6
1.4 Workmanship and FFS 7
The sponsoring editor for this book was Kenneth P. McCombs, the editing 1.5 FFS in Construction Codes 9
supervisor was David E. Fogarty, an.d the prod.uction supervisor was Pamela A. 1.6 The Fitness-for-Service Step 10
Pelton. It was set in Century Schoolbook by Wayne A. Palmer and Pat Caru.~o
1.7 Three Criticai Questions 10
o{McGraw-Hill Professional's Hightstown, N.J, composition unit. The art
director for the cover was Handel Low. 1.8 Maintenance Strategy 11
1.9 Pressure Boundary lntegrity 11
Printed and bound by RR Donnelley. 1.10 The Five Disciplines 13
1.11 Regulatory Perspectiva 14
~ This book was printed on recycled, acid-free paper containing a
1.12 Codes, Standards, and Guides 15
\!l:,l minimum of 50% recycled, de-inked fiber. 1.13
1.14
Cum L.aude
Technical Basis
16
18
1.15 Response Time 19
McGraw-Hill books are available at special quantity discounts to use as 1.16 Summary 21
premiums and sales promotions, or for use in corporate training programs. For References 21
more information, p\ease write to the Director of Special Sales, McGraw-Hill
Professional, '1\vo Penn Plaza, New York, NY 10121-2298. Or contact your \o<:al
bookstore. Chapter 2. Materiais 25
2.1 Demand and Capacity 25
2.2 Material Groups 25
Information contained in this work has been obtained by The McG•·aw-Hill 2.3 Ferrous Metais 26
Comparues, Inc. I"McGraw-Hill"l from sources believed to be reíiable. However,
neither McGraw-Hill nor its authors guarantee the accuracy or completeness of 2.4 Nonferrous Metais 31
any infonnation pu_blished herein and neither McGraw-Hill nor its authors 2.5 Nonmetallic Materiais 33
shall be responsible for any errors, omissions, or damages arising out of use of 2.6 Basis for Material Selections 34
this information. This work is published with the w1derstanding that McGraw·
Hill and its authors are supplying information but are not attempting to render 2.7 Mechanical Properties Overview 34
engineering or other professional services. If such services are required, the 2.8 lrlow to Achieve Desired Properties 35
assistance of an appropriate professional should be sought. 2.9 Phase Diagram of Carbon Steel 36
2.10 Heat Treatment 39
2.11 Benefits of Postweld HeatTreatment 40
v
vi Contents Contents vil
2.12 Types of Heat Treatment 41 3.31 Fatigue 97

2.13 Shop and Field Heat Treatment 42 3.32 The ASME Boller and Pressure Vessel C ode Fatigue Method 100
2.14 The Larson-Miller Parameter 45 3.33 The Markl Fatigue Method 104
2.15 Heat and Lot 45 3.34 Example of the Markl Method In Vlbratlon 105
2.16 The Three Strength Parameters 47 3.35 The Fracture Mechanlcs Fatigue Method 106
2.17 Allowable Stress 49 3.36 The AWS.AASHTO Fatigue Method 108
2.18 Obtalnlng Strength Propertles of Operating Equipment 50 3.37 Fatigue Testlng 109
2.19 Factors Atfectlng Strength Propertles 52 3.38 ASME Stress Classlflcatlon along a Llne 110
2.20 Ductlllty 52 3.39 Externai Pressure 112
2.21 Ductlle Fracture 53 References 114
2.22 Brittle Fracture 55
2.23 Toughness 58 Chapter 4. Fabrlcatlon 117
2.24 Charpy Toughness 59
2.25 Fracture Toughness 60 4.1 Fabrlcation and Construction Flaws 117
2.26 Toughness Exemptlon Curve 62 4.2 Base Metal Defects 118
2.27 Hardness 62 4.3 Fabrlcation Flaws 120
References 64 4.4 Welding Techniques 121
4.5 Carbon Equivalent 124
4.6 Weld Quallty 125
Chapter 3. Deslgn 67 4.7 Welding in Service 126
3.1 Basic Deslgn and Detalled Deslgn 67 4.8 Pressure or Leak Testlng-How? 130
3.2 Deslgn Codes 67 4.9 Pressure or Leak Testlng-Why? 131
3.3 Deslgn Mlnlmum Wall t m 1n 69 4.10 Pressure or Leak Testlng-Cautlons 133
3.4 Future Corroslon Allowance FCA 69 4.11 Test Pressure for Tanks 133
3.5 Loads, Stresses, and Strains 71 4.12 Test Pressure for Pressure Vessels 134
3.6 Applled Loads and Residual Stresses 73 4.13 Test Pressure for Power Piplng 135
3.7 General Stresses 74 4.14 Test Pressure for Process Plplng 135
3.8 Example: Bendlng Stress 74 4.15 Test Pressure for Llquld Plpellnes 136
3.9 Pressure Stress 76 4.16 Test Pressure for Gas Plpellnes 136
3.10 Pressure Stress Example 76 4.17 Mlll and Handllng Flaws 137
3.11 Wall Thlckness Selectlon n 4.18 Fleld Weld Flaws 137
3.12 Fossll Power Plant Example 78 4.19 Weld Slze 143
3.13 8utt·Welded Flttings 79 4.20 Residual Stress 143
3.14 Flanges 79 4.21 Measuring Residual Stresses 145
3.15 Socket and Threaded Flttings 79 4.22 Calculating Residual Stresses 147
3.16 Specialty Flttlngs and Components 80 4.23 Mechanical Jolnt Flaws 149
3.17 Vessel Example 80 References 149
3.18 Deslgn Principies 82
3.19 Deslgn Pressure 83 Chapter 5. Oegrada1ion 153
3.20 Vessel Cyllndrlcal Shell 84
3.21 Spherlcal or Hemlspherical Head 86 5.1 Corroslon 153
3.22 Elllptlcal Head 87 5.2 The Corroslon Englneer's Perspectiva 153
3.23 Torispherlcal Head 89 5.3 The Facllity Englneer's Perspectlve 154
3.24 Flat Head 89 5.4 Damage 156
3.25 Comparlson 90 5.5 Degradatlon and Fltness-for-Service 156
3.26 Plant Plplng- ASME 831.3 91 5.6 Understandlng Wall Thlnnlng Mechanisms 156
3.27 Plant Plplng Moment Stress 93 5.7 The Electrochemlcal Cell 157
3.28 Applled Forces 95 5.8 The Single Metal Electrochemlcal Cell 159
3.29 Liquid Pipellnes- ASME 831.4 96 5.9 The Galvanlc Cell 160
3.30 Gas Plpellnes 97 5.10 Concentration Cell 161
vlll Contents Contents lx
5.11 Slze Effect 163 6.5 How? 214

5.12 Parameters Affecting Corroslon Rate 163 6.6 When? 215
5.13 Predicting Corroslon Rate---ls lt Linear? 165 6.7 Rlsk·Based lnspectloll-What ls Rlsk? 216
5.14 Predicting Corroslon Rate-Time in Service 166 6.8 A Number or a Matrix 216
5.15 Deposlts and Tuberculatlon 166 6.9 The Objectlve of RBI 218
5.16 General Corrosion 167 6.10 Necessary and Sufflcient 218
5.17 Galvanic Corrosion 169 6.11 ls RBI a Cost-Saving? 219
5.18 Pltting 169 6.12 Qualitativa or Quantitativa RBI 219
5.19 Crevlce Corrosion 170 6.13 RBI: A Seven-Step Process 220
5.20 Corroslon under lnsulatlon 173 6.14 Qualitativa RBI 222
5.21 Liquid·Line Corrosion 175 6.15 Example of Quatltatlve RBI-Steam Systems 223
5.22 Microbial-lnduced Corroslon 175 6.16 Semlquantltative RBI-Likelihood 226
5.23 MIC Preventlon 177 6.17 Semlquantltative RBI-Consequence 227
5.24 MIC Mitigation 177 6.18 Fully Quantitativa RBI-llkelihood 228
5.25 Carbon Dioxlde Corrosion 178 6.19 General Metal Loss Example 229
5.26 Erosion 179 6.20 Llmit State Functlon for a Crack 230
5.27 Cavitation 179 6.21 Crack Example 230
5.28 Vapor-Liquid Eroslon 181 6.22 Llkellhood Reduction Options 231
5.29 Erosion in Gas-Liquld Servlce 181 6.23 Correctlon for Reliability 232
5.30 Llquid Pipelines 182 6.24 Fully Quantitative Consequence 233
5.31 Llquid-Sand Pipellnes 182 6.25 Advantages of Applying RBI 237
5.32 Eroslon-Corrosion 182 6.26 Cautions When Applying RBI 237
5.33 Environmentai-Asslsted Cracking Mechanisms 184 6.27 integrity Programs for Plpellnes 238
5.34 Corrosion Fatigue 186 6.28 Overview of lnspection Technlques for Tanks, Vessels, and Pipes 239
5.35 Sensitized Stainless Steel 186 6.29 Visual Examination (VT) 239
5.36 Sour Corrosion 188 6.30 Magnetlc Partlcle Testlng (MT) 240
5.37 Bllsters and Cracks 190 6.31 Llquid Penetrant Testlng (PT) 241
5.38 Hlgh-Temperature Corroslon 192 6.32 Radlographlc Testing (RT) 242
5.39 Measurlng Corroslon and Corrosion Rates 194 6.33 Ultrasonlc Testlng (UT) 243
5.40 Coating 196 6.34 Long-Range Gulded Wave Ultrasonlc lnspectlon 246
5.41 Common Coatings 196 6.35 Eddy Current Testing 246
5.42 Selection 197 6.36 Magnetlc Flux Leakage 246
5.43 Surface Preparation 198 6.37 Acoustlc Emisslon Testing (AE) 248
5.44 WrapTape 199 6.38 Plg lnspectlons of Pipelines 248
5.45 Epoxy 200 6.39 Leak Detection Systems 249
5.46 Coai Tar Enamel 202 6.40 Direct Assessment 249
5.47 Heat-Shrinkable Sleeves 203 References 249
5.48 Multllayer Coating 203
5.49 Coatlng Performance 204 253
Chapter 7. Thlnnlng
5.50 Coatlng Quality Control 205
5.51 Comparlson 207 7.1 Three Categorias of Wall Thlnning 253
5.52 Practical Challenges 207 7.2 Leak or Break 254
References 7.3 When ls Corroslon Considered General? 254
208
7.4 Principies of Evaluation for GML 257
7.5 Llmltatlons 258
Chapter 6. lnspection 213 Bucktlng 258
7.6
6.1 Principies of lnspectlon 213 7.7 Ultrasonlc Grld 258
6.2 Why? 213 7.8 Storage Tank Example 259
6.3 What? 214 7.9 API 653 Tank Thickness 262
6.4 Where? 214 7.10 Tank Fltness-for-Service-Level 1 263
X Contents Contents xl
7.11 Resolutlon 265 9.10 Foundatlon of Fracture Assessment, the FAD 317
7.12 Tank Fltness-for-Service-Level 2 265 9.11 The 15 Steps of Crack Analysls 319
7.13 Power Plant Plpe Ruptura 266 9.12 Vessel Example 320
7.14 Power Plpe Fltness-for-Service-Level 1 266 9.13 Margln to Fallure 323
7.15 Process Plpe Fltness-for-Service-Level 1 269 9.14 Leak Through Cracks 325
7.16 ASME 831 G for Plpellnes: What ls lt? 270 9.15 Appllcatlon of Fracture Mechanlcs to Fatigue 326
7.17 8asls of ASME 831G 271 References 327
7.18 Deratlng a Plpellne 273
7.19 831G Example for Gas Pipeline 274 Chapter 10. Creep Damage 329
7.20 Modified 831G 274
7.21 The RSTRENG• Method 276 10.1 What ls Creep? 329
7.22 The Remainlng Strength Factor in B31G 276 10.2 Hlgh-Temperature Corroslon 331
7.23 Steam Condensate 277 10.3 The Difficultles of Creep Analysis and Predictions 332
7.24 The ASME VIII, Dlv.1, App. 32 Method 279 10.4 Short- and Long-Term Overheating 335
7.25 The ASME XI Code Case N-480 Method 281 10.5 Creep Assessment Methods 336
7.26 Wldespread Pittlng 281 10.6 ASME 111 NH Method 337
7.27 Locallzed Pitting 282 10.7 Operatlng Loads 337
7.28 Example Plttlng in Pipellne 283 10.8 Time-lndependent Material Properties 338
7.29 Slmple Crlterion 286 10.9 Time-Dependent Material Propertles 338
References 287 10.10 Creep Life Analysls 340
10.11 Qualificatlon 342
10.12 API 530 Creep Assessment 343
Chapter 8. Geometric Defects 289 345
10.13 Nondestructlve Assessment
8.1 lntegrlty of Geometric Defects 289 10.14 High-Temperature 831.3 Plpe Appllcatlon 346
8.2 Assessment Steps 289 10.15 Draft Method of API 579 Level1 347
8.3 Distorted Tank Bottom Example 291 10.16 Life Fractlon Analysls 347
8.4 Accidentally Bent Rlser Example 292 10.17 Thlnned Wall Remalnlng Llfe 348
8.5 Dents In Plpellnes 294 10.18 Metallographlc Llfe Assessment 349
8.6 Dents wlth Gouges 296 Aeferences 350
8.7 Wrinkles and Buckles 298
8.8 Mlld Ripples 300 351
Chapter 11. Overload
8.9 811sters 300
8.10 Fltness-for-Service of Equipment with 81isters 300 11.1 Overloads in Practice 351
8.11 Assessment of Weld Misallgnment and Shell Distortions 302 11.2 Overpressure Allowance 351
8.12 Fltness-for-Service Assessment of Peaking 302 11.3 Overpressure beyond Allowance 352
8.13 Deformed and Repaired Vessel Explosion 305 11.4 Key Conslderations for Overpressure 354
8.14 Defects 8eyond Assessment 307 11.5 Waterhammer Overload 355
References 307 11.6 Bolted Jolnt Fallure 357
11.7 The Bullet Plg 360
11.8 Detonatlons and Deflagratlons 360
Chapter 9. Cracks 309 361
11.9 Exploslon Pressures
9.1 Crackllke Flaws 309 11.10 Exploslon Damage 363
9.2 Crack Stablllty 310 11.11 Example-Deflagration In Plpe 364
9.3 Stress lntensity 311 11.12 Material Strength at Hlgh Straln Rates 364
9.4 Example-Crack in Pipellne 312 11.13 Exploslve Rupture and Fragmentation 365
9.5 Stress lntenslty Solutions 314 11.14 Effect of Externai Exploslons 368
9.6 Fracture Toughness 314 11.15 Natural Phenomena Hazards 369
9.7 Weld Residual Stresses 315 11.16 Fltness-for-Servlce by Plastlc or Collapse Analysis 372
9.8 Ligament Reference Stress 315 11.17 Bending Fallure 373
9.9 Flow Stress 317 References 374
xii Contents
Chapter 12. Failure Analysis 377 Preface

12.1 Fallure Mode and Effects 377
12.2 Root Cause Fallure Analysis 378
12.3 Fallure Analysis Tools 380
12.4 Leak·before-Break (LBB) 382
12.5 Stored Energy Assocíated with Flashing Liquids 384
12.6 Hydrotest Failure 384 This book is an exploration offailure in tanks, vessels, and pipes. This is
12.7 Gas or Liquid Contents 385 done from an engineer's perspective, which means that it has to be usable
12.8 The Tank Top Example 386 to make run-or-repair decisions, to decide whether the component is still
12.9 Tanks wlth Frangible Roof Design 386 "fit for service," or whether it should be repaired or replaced.
12.10 Stored Energy 388 Fitness-for-service is the understanding ofwhy, when, and how tanks,
12.11 Leak-before-Break Using the Failure Assessment Diagram 389
vessels, and pipes fail. The answer is at the crossroads of mechanical
References 389
design and analysis, metallurgy, welding, corrosion, inspection, anda
good grasp of operations. These disciplines have to merge into a succinct
Chapter 13. Repairs 393 and clear answer, preferably a single number: how much longer can the
13.1 Repair Work Package 393 component operate safely. In answering thls question, the engineer will
13.2 Postconstructlon Codes and Standards 393 have to rely on experience and science, and will have to reach a quanti-
13.3 Temporary or Permanent Repair? 395 tntive answer in the form of time to failure or margin to failure.
13.4 Safety 395 This book is a modest addition to the many documents that address
13.5 Regulatory Requirements 395
the interesting question of fitness-for-service. Foremost among publi-
13.6 Common Conslderatlons for Materiais 396
cations dealing with the topic is the American Petroleum Institute's
13.7 Common Considerations for Design 39~
13.8 Common Considerations for Fabrication-Welding 397
Recommended Practice API 579 Fitness for Service, a document which
13.9 Controlled Deposition Welding 398 through its elegance, completeness, and practical value remains un-
13.1 o Postconstruction Standards for Controlled Deposition 401 matched. The reader is urged to apply the API practice and other codes
13.11 Common Considerations for Fabrication-Nonwelding 402 and standards referenced in the text, as they are the authoritative refer-
13.12 Common Considerations for Examination 4()2 ences to be used in making fitness-for-service and run-or-repair decisions.
13.13 Common Considerations forTesting 4()3
13.14 Common Conslderatlons for Quality Control 405 George Antaki, PE
13.15 Replacement 405 Aiken, South Carolina
13.16 Flush Patch Repair 406
13.17 Example of Flush Patch Repalr 407
13.18 Flaw Excavation 407
13.19 Example of Flaw Excavation Repair 409
13.20 Weld Overlay 409
13.21 Full Enclrclement Welded Sleeve 411
13.22 Welded Leak Box 413
13.23 Fillet-Welded Patch 414
13.24 Mechanical Clamp 417
13.25 lnserted Liner 418
13.26 Pipe Splitting 419
13.27 Sacrificial Component 419
13.28 Nonmetallic Wrap 419
References 422
Appendix. WRC Bulletins 425
lndex 451
XIII
Chapter
1
Principies
1.1 What ls Fitness-for-Service?

Fitness-for-service (FFS) is a set of quantitative methods used to deter-
mine the integrity and remaining life of degraded components, and to
make run-or-repair decisions. FFS is also referred to as fitness-for-pur-
pose or mechanical integrity. It applies to storage tank.s, pressure ves-
sels, boilers, piping systems, and pipelines. This type of equipment is
sometimes referred to as static or fixed equipment, in contrast to equip-
ment with moving parts such as pumps and compressors referred to as
dynamic or rotating equipment. Fitness-for-service, as addressed in
this book, covers static (fixed) equipment, and the pressure boundary of
dynamic (rotating) equipment. FFS does not address the operation of
active equipment such as pumps, fans, or compressors.
1.2 FFS and Conduct of Operations

Fitness-for-service is one step in the overall process of equipment
integrity. Health and longevity of mechanical equipment is similar to
that of humans; it depends on (1) its genes (how the equipment was
designed, how the materiais were procured, and how the equipment
was constructed); (2) its lifestyle (how the equipment is operated); and
(3) the quality ofits checkups, especially in old age or when abnormal
symptoms appear (how the equipment is maintained and inspected,
and how its fitness-for-service is periodícally evaluated). These aspects
are illustrated in Fig. 1.1. Steps 1 to 3 address design and construction
(the "genes"), Step 4 is operation (the "lifestyle"), and Step 5 address-
es fitness-for-service (the "checkups").
2 Chapter One Principies 3
• Is coating, cathodic protection, or lining required?

1 1.1 - Selection • Are material selection assumptions incorporated into basic
-+ +-- system design and monitoring?
Materiais 1.2 - Procurement Quality
• Are corrosion inhibitors required? Are they input to basic
.. system design?
• Is a corrosion rate established?
• Is a design-replacement life developed based on the corrosion
2 2.1 - Basic System Design
Design 1--. 2.2 - Detailed Design
f4- rate?
• Is there a need for corrosion monitoring in-service?
... • Is there evidence of corrosion not accounted for in design?
(faiiure analysis)
3.1 - Shop Fabrication & • Is the corrosion rate larger than the design corrosion. allow-
3 ance? (failure analysis)
Construction f-+ Field Erection f4-
3.2- lnspection & Testing 1.2. Materiais Procurement Quaiity
.. • Are the materiais completely and cleariy specified?
• Is the material supp1y chain understood and trustworthy?
4.1 - I&C, Procedures, & • Are quality control measures in piace: manufacture, distrib-
4 ution, receipt, and stores?
-+ Training ~
Operation • Is material traceable? (positive material identification PMI).
4.2 - Emergency Response
• Are dimensions and finishes per spec.?
+ • Is the material chemical composition per spec.?
• Is the microstructure as expected?
5.1 - Risk-Based Monitoring
5 • Is strength (yield, ultimate, elongation at rupture) per spec.
Maintenance f-+ & Fitness-for-Service f4-
5.2 - Management of Change mínima?
• Is toughness (CVN, K 1c, etc.) unusually low?
Figure 1.1 Conduct of operations. • Is documentation complete, correct, clear, and retrievable?
2. Design
2.1. Basic System Design
• Will the basic process work as expected (proof of design)?
The logic of Fig. 1.1 is the guiding principie that applies to all activ-
• Is there a potential for exothermic or explosive reaction?
ities, from large projects to simple repair packages. Fitness-for-
• Is the safety logic sufficient and operational?
Service, step 5.1 in Fig. 1.1, is a fundamental component, in the overall
• Can the process deliver the throughput as expected?
conduct of operations, a strategy meant to achieve safe and profitable
• Can the throughput be readily controlled to stay within limits?
?peration. Figure 1.1 can be used as a planning tool, looking forward
• Are instruments and controls sufficient and operational?
m planning engineering activities, oras an investigative tool, looking
• Are instruments and controls properly calibrated?
back at the root cause of a failure. For practical use, Fig. 1.1 can be
• Is overpressure protection adequate?
expanded in the form of lines of inquiry.
• Is fire protection adequate?
1. Materiais • Will alarms work as planned? Are they sufficient?
1.1. Materiais Selection • Is the safety analysis (safety case) consistent with ali aspects
• Is the material compatible with the process chemistry? of basic design?
• Is the material compatible with the operating pressure, tem- • Are there system descriptions and emergency respon.se pro-
perature, and flow rate? tocais and procedures?
• Are degradation mechanisms correctly identified? • Is documentation complete, correct, clear, and retrievable?
2.2. Detailed Design • Is preoperational testing and turnover to operations satis-

• Are operating and design conditions well defined and consis- factory?
tent with basic design? • Is documentation complete, correct, clear, and retrievable?
• Are design codes and standards specified (civil, mechanical, 4. Operation
electrical, I&C, tire protection)? 4.1. Instrumentation and Controls, Procedures, and Training
• Are there regulatory or contractual requirements beyond • Are there controls of the product stream (on the inside of
codes and standards? equipment)?
• Are design processes, interfaces, and responsibilities well • Are there controls of the environment (on the outside of
delineated? equipment)?
• Are designers competent? • Are instruments and controls sufficient for safe operation
• Is the system protected against overpressure? (pressure, temperature, flow)?
• Is layout and support adequate? • Are the operating procedures adequate? Are they followed by
• Is flexibility and strength sufficient for the range of operat- operators?
ing temperatures? • Are operating records and logs complete, clear, and retrievable?
• Is thermal shock or fatigue addressed in design? • Are shift turnover processes adequate?
• Will there be unusually large vibration? • Are operators trained and do they share lessons learned,
• Will thermohydraulic transients occur? within company and industrywide?
• Are natural hazards a part of design (soil settlement, high 4.2. Emergency Response
wind, earthquake, etc.)? • Are operators trained and drilled to follow emergency proce-
• I s documentation complete, correct, clear, and retrievable? dures?
3. Construction • Do operators understand the system function, normal and
3.1. Shop Fabrication and Field Erection abnormal?
• Are design, fabrication, and erection drawings complete and • Do operators understand the safety basis and operating lim-
clear? its of systems?
• Have there been a constructability review and buy-in engi- • Are operators required to place safety first, and empowered
neering-construction? to shut down if necessary?
• Are receipt inspection 1;111.d stores control processes in place? • Are local plant and general community notification protocols
• Is shop fabrication quatt.y acceptable? in-place?
• Is field erection quality acceptable? • Are emergency responders on call, trained, and ready?
• Is welding quality within code? 5. Maintenance
• Are there abnormal loads introduced during handling, align- 5.1. Risk-Based Monitoring and Fitness-for-Service
ing, and erection? • Are systems and components classified based on risk (likeli-
• Is there mechanical joining per vendor requirements and codes? hood and consequence of failure)?
• Are flanged joints properly assembled and bolted? • Is there a risk-based monitoring (inspection) program?
• Is construction per design? Have deviations been accepted by • Are high-risk components periodically inspected, monitored
engineering? (predictive maintenance)?
• Is tbere a fmal as-built review? • Are low-risk components unnecessarily inspected (run-to-
• Is documentation complete, correct, clear, and retrievable? failure, corrective maintenance)?
3.2. Inspection and Testing • Are the right spots inspected, with the right technique and
• Is nondestructive examination (NDE) per code? the right interval?
• Are NDE examiners certified and supervised? • Are inspectors certified? Are they sufficiently independent?
• Is leak testing conducted in accordance with construction • Are inspection results analyzed and trended in a timely
code? Is it adequate? manner?
• Is fitness-for-service based on competent engineering, or on]y

on "judgment"?
• Are reliability data from maintenance fed back to engineer-
ing and operations?
• Is maintenance history saved and linked to the equipment
through a database?
• Is documentation complete, correct, clear, and retrievable?
5.2. Management of Change (MOC) Figure 1.2 Externally, pipe weld
• Is there a formal change control process? appears acceptable.
• For criticai processes, does MOC conform to regulations?
• Does the MOC program include a competent job hazard
analysis?
• Are change packages prereviewed and signed-off by engi-
neering, safety and operations?
• Is there a safe lockout-tagout process?
• Are changes designed, constructed, and tested to codes and
standards?
• When change is due to failure, is there a root-cause failure
analysis process?
• Are changes implemented by qualified personnel (welding,
inspection, testing)?
• Are completed changes reviewed for as-built conformance to
design?
• Are changes reflected in plant drawings and procedures?
• Are completed change packages signed-off by engineering,
safety, and operations prior to service?
• Is documentation complete, correct, clear, and retrievable? Figure 1.3 Intemally, weld is of poor quality.
1.3 Fitness-for-Service of Old

and New Equipment Altematively, and if agreed by ali parties, including the owner, the juris-
dictional authority, and the underwriting agency, the owner may choose
A new structure, system, or component (SSC) correctly designed, fabri- to evaluate the defect for fitness-for-service, to assess whether it may be
cated, installed, inspected, and tested is of course fit-for-service. Fitness- safely left as is. In this case, of course, it also becomes necessary to search
for-service is therefore of primary interest for SSCs that have been in for and find any other such defects in the system, and to evaluate all the
service and have undergone some degree of degradation. defects for fitness-for-service, and repair them if necessary.
But fitness-for-service can also prove valuable for new equipment, par-
ticularly if a defect is discovered late in the constructíon process, when
repair is no longer a viable option. For example, a new component may 1.4 Workmanship and FFS
exhibit a base material or a fabrication defect that exceeds the limits of The best industrial materiais, welds, and construction contain initial
the construction code; in the case of Fig. 1.2, a weld in a newly installed flaws. These flaws are limited by construction codes to a certain size.
piping system looked quite good from the outsíde, but a final field inspec- For example, the pipeline welding standard permits slag inclusions in
tion unveiled severa! weld defects (Fig. 1.3). The poor workmanship had a weld, provided their aggregate length does not exceed 2 in. 1 Such
gone unnoticed during construction. In this case, the construction code limits on defect size in a construction code are not based on a quanti-
would require weld defects of this magnitude to be repaired. tative analysis of the safety margin of the weld flaw, nor are they
based on fitness-for-service of a component with the defect; instead,

these acceptance limits are traditional and proven workmanship qual-
ity standards. A good welder, with a good welding procedure, should be
able to make a weld with slag inclusions shorter than 2 in.
Construction code workmanship standards "are based on empírica]
criteria for workmanship and place primary importance on imperfec-
tion length. Such criteria have provided an excellent record of reliabi1-
ity in pipeline service for many years. "1 Figure 1.5 Incomplete penetration weld.
Defects that exceed the construction code Iimits will not necessarily
fail in service. For example, the slag inclusions in Fig. 1.4 exceed the
2-in limit of the construction code, but the component operated with-
out problem for decades, until its retirement. But the workmanship
standards of construction codes are (a) imposed by contract and, in
some cases, by regulation, and (b) a proof of competency; for these rea-
sons they must be met, unless fitness-for-service is an option formally
accepted by all parties.
Figures 1.5 and 1.6 illustrate another example of the difference
between workmanship standards and fitness-for-service. Two 24-in
carbon steel pipe spools, %-in thick, are welded together by a circum-
ferential butt weld that, on purpose, has a %2-in incomplete penetra-
tion (Fig. 1.5). The welded assembly was filled with water and the
pressure steadily increased until rupture occurred at 2250 psi. As
evidenced in Fig. 1.6, the rupture occurred as a longitudinal ductile
Figure 1.6 Burst test with

incomplete penetration weld.2
fishmouth burst in the base material; the rupture did not occur along
the flawed circumferential weld. The incomplete weld was still fit-
for-service.
1.5 FFS in Construction Codes

Few construction codes permit the use of FFS methods to judge and
Figure 1.4 These weld inclusions did not jeopardize component accept fabrication flaws. Instead, most construction codes rely on
operation. workmanship standards, and if a defect exceeds the standard then the
part is repaired, prior to service. The pipeline welding standard API the component is criticai for financiai or safety reasons. When degra-
1104 is one of the few exceptions that recognize fitness-for-service as dation is suspected or detected, plant staff is called upon to answer
an option for new weld defects: ''Fitness-for-purpose criteria provide these criticai questions:
more generous criteria allowable imperfection sizes, but only when
additional procedure qualification tests, stress analyses, and inspec- 1. What caused the degradation? How fast is it progressing?
tions are performed."1 The same allowance for "more generous ente- 2. What is the margin to failure? How long can we keep operating as is?
ria" is not recognized in most other construction codes, among them
3. Ifleft unchecked, how will the component fail? Will it be a leak, a rup-
Section VIII of the ASME Boiler an d Pressure Vessel Code.3
In fact, accepting fitness-for-service as a policy in new construction ture, or structural collapse?
would be prohibitive because construction would be riddled with The answer to the first question resides in operating experience,
delays and cost overruns to analyze defects, gain approvals, and inves- materiais, corrosion, and construction knowledge. The answer to the
tigate the extent of poor workmanship. second question is what fitness-for-service is all about. The answer to
the third question is based on a combination of experience, and stress
1.6 The Fitness-for-Service Step and fracture analyses. It is hoped that the reader will find in this book
useful information in responding to these three criticai questions.
Step 5.1 of Fig. 1.1 is labeled Risk-Based Monitoring and Fitness-
for-ServicP. RiRk-h::~RPrl monitor in g is m ore commonly referred to as
risk-based inspection (RBI) and consists of (1) inspection planning, 1.8 Maintenance Strategy
(2) conduct of inspection, (3) assessment of inspection results, and Another way oflooking at fitness-for-service is to see that it isto stat-
(4) actions and feedback. ic equipment (tanks, vessels, and piping) what predictive maintenance
In inspection planning, the systems are ranked on t he basis of r:isk isto rotating (dynamic) equipment. It is the third of the three mainte-
<likelihood and consequence of failure). The h:igh-risk systems are pri- nance strategies:
oritized for inspection, defining what to inspect. Inspection planning
also includes decisions regarding where, how, and when to inspect. 1. Correctiue maintenance (CMJ. Run to failure. This is an acceptable
These aspects of inspection planning are addressed in Chap. 6. strategy for low-risk equipment.
Following inspection planning, the inspections will take place.
Inspection techniques include nondestructive testing, or pressure and 2. Preuentiue mamtenance (PMJ. lnspect at fixed intervals, and accept
leak testing. The application of th ese t echniques in support of FFS or reject on the basis of construction code. This approach is difficult
assessments is also discussed in Chap. 6. to justify for static equipment, as it may lead to either overinspect-
The next step, the assessment of inspection r esults, is the focus of ing or underinspecting.
th:is book. The outcome ofthis step isto decide whether the equipment 3. Predictiue maintenance (PdM). Inspect and evaluate on the basis of
is fit-for-service, and-if yes-for how long. engineering analysis, trending, and margins. This is the right strat-
From there, if significant degradation is observed, there is a feedback egy for high-risk equipment. It does require mor e technical exper-
loop to changes in materiais, design, fabrication, operation, or inspec- tise in setting inspection intervals, techniques, locations, and in
tion. For example, if the equipment is fit for service for another ten trending and analyzing the inspection results.
years, the inspection interval may be modified to inspect the equipment
at half the remaining life, in this case five years. If, over time, there is
1.9 Pressure Boundary lntegrity
no sign of degradation, the inspection plan may be revised to eliminate
the inspection altogether, or change the inspection location or technique. When embarking on an integrity program, it is essential to define the
program objective. Fitness-for-service, as commonly applied in the
power, pipeline, and process industries, focuses on the integrity of
1.7 Three Criticai Questions
the base material itself and its welds. lt does not address the oper-
In practice, fitness-for-service becomes necessary when (1) degrada- ability of a system (whether pumps, compressors, or valves will per-
tion is suspected or observed in a component, and (2) the integrity of forro their function), and-in many cases-fitness-for-service does not
address the leak tightness of mechanical joints (threaded fittings,

flange joints, specialty fittings, expansion joints, etc.). This is the nar-
row but common scope of fitness-for-service: integrity of the pressure
boundary's base metal and welds.
An operating company may choose to include in fitness-for-service
the leak tightness of mechanical joints and the general condition of
equipment supports. In this case, the volumetric and surface exami-
nations of base material and welds would be supplemented by visual
inspections ofthe general condition ofthe equipment and its supports,
and leak tightness check of joints.
For example, in Fig. 1.7, addressing the general condition of the
pump and its supports may or may not be part of a company's fitness-
for-service program.
In Fig. 1.8, gas leakage through a flange gasket is discovered by
spraying bubble solution at the flange joint. Leak tightness o f mechan-
ical joints may or may not be part of a company's fitness-for-service
program. Figure 1.6 Out-lettlutge of ttir ueLecku uy uuuule :sululiun.
The broadest approach to mechanical integrity will address ali com-
ponents in a system, the system's operation, its maintenance history,
and its reliability. Such reviews are captured and trended in System Health Reports that track the reliability of structures, mechanical and
electrical equipment, and instrumentation and control systems.
1.1 O The Five Disciplines

Fitness-for-service, the assessment of the remaining life of a compo-
nent, has to be implemented by a group of persons, or by a person (if
such a person does exist), with expertise in tive disciplines: materiais
and corrosion, stress analysis and codes and standards, fabrication
and welding, inspection, and operation ofthe system under evaluation.
The responsibility of each discipline is as follows:
• Materiais and corrosion. To understand the type of degradation (is it
general wall thinning, local attack, cracking, etc.?) and what is caus-
ing the degradation, to help select inspection locations, and to deter-
mine the corrosion rate.
• Stress analysis and codes and standards. To predict the component's
current and future integrity, to help select inspection locations based
on stress margins, and to perform the fitness-for-service assessment
(which is primarily based on stress and fracture analysis).
• Fabrication and welding. To understand the quality of construction,
the significance of flaws, to have a feel for residual stresses (criticai
in the evaluation of cracklike defects), and to differentiate between
Figure 1.7 General corrosion of pump support. initial construction defects and service-induced flaws.
• Inspection. To perform the inspections, to help select inspection loca- Periodic inspections and :fitness-for-service are self-imposed by most
tions, to plan the inspection, to interface with safety personnel for plant operators in the chemical process, re:fining, and petrochemical
access, permits, protective gear, and the like, to be experienced and industries. Applicable standards include API 510, 570, 572, and 574.
certified in inspection techniques, their accuracy and applicability, Periodic inspections and :fitness-for service are imposed in many
and recognize the degradation mechanisms at play. states by state regulations, particularly for boilers. Oversight is pro-
• Operation. To help understand the normal operating conditions, the vided through a state inspection office, and the guiding document is
past ah normal transients, and future operating plans. Oftentimes, the the National Board oflnspectors Code (NBIC).
operator is the person too busy running the facility to have time to sit
down and contribute to the fitness-for-service effort. Yet, the opera- 1.12 Codes, Standards, and Guides
tor's input is crucial in at least understanding if the process has been
run as expected, and when and how the process stream has changed, In the hierarchy of codes and standards, fitness-for-service standards are
and to help flag upsets and unusual occurrences that could explain an extension of the older and better-established design and construction
why a certain degradation mechanism appeared at a given time. codes and standards (Fig. 1.9). The interest in fitness-for-service carried
from the aeronautical industry into the process, power, and pipeline
industries. Several standards were developed, often addressing a partic-
1.11 Regulatory Perspective ular industry, such as pipelines or nuclear power plants. These FFS stan-
Periodic inspections and fitness-for-service assessments are mandato- rlards include the following:
ry in the nuclear power industry, through the Code of Federal
• API 1104, Welding of Pipelines and Related Facilities, Appendix A,
Regulations 10CFR50.4 Implementation is throughASME XI5 and reg-
Alternative Acceptance Standards for Girth Welds
ulatory oversight by the U.S. Nuclear Regulatory Commission (NRC).
Periodic inspections and integrity management are mandatory in the • ASME B31G, Manual for Determining the Remaining Strength of
oil and gas pipeline industry, through the Code of Federal Regulations Corroded Pipelines, a Supplement to ASME B31 Code for Pressure
49CFR.6 Implementation is through the regulation itself, ASME Piping
B31.4,7 B31.8,8 and B31.8S,9 as well as API recommended practices. • Canadian Standard Association CSA Z662, Oil and Gas Pipeline
Regulatory oversight is by the U.S. Department of Transportation's Systems
Office of Pipeline Safety (OPS).
• DVS Guidelines 2401, Fract ure-Mechanical Eval uation of Faults in
Periodic inspections and :fitness-for-service are mandatory in the
Welded Joints; Fundamentais and Procedure, Fracture-Mechanical
chemical industry when the process involves flammable or toxic materi-
Evaluation of Faults in Welded Joints; Practical Application,
ais above a certain threshold quantity. Explicit requirements are con- Germany
tained in the Office of Safety and Health Administration (08HA)
regulation 29CFR 1910.11910 under the heading, "Mechanical Integrity." • European Pipeline Research Group, EPRG Guidelines
They include the following: • SINTAP, Structural lntegrity Assessment Procedures for European
lndustry
• Identification and characterization of equipment and instrumenta-
tion • PrEN 13445-3, 1998 Fatigue Verification ofWelded Joints, European
Standard
• De:finition of inspections and tests, and their frequency, based on
degradation rate • British Standard Institute BS7910, Guide on Methods for Assessing
the Acceptability of Flaws in Metallic Structures, an extension and
• Development of inspection acceptance criteria, consistent with codes update ofPD-6493, Guidance on Some Methods for the Derivation of
and standards Acceptance Leveis for Defects in Fusion Welded Joints
• Documentation of inspections and assessments • Australian Standard AS 2885, Pipelines-Gas and Liquid Petroleum
• Management of change program (SAA Pipeline Code)
• Incident investigation program • ASME B&PV Code, Section XI, In-Service lnspection
• SQNFoU A Procedure for Safety Assessment of Components with

Cracks- Handbook, Sweden Design and EitDflSS-fQ[-
Equipment Class
Construction ~
• CEGB Report RIHIR6, Assessment of the Integrity of Structures i!!UI B!Uli!irll
Containing Defects
• Canadian Standard CAN/CSA-N285.4, Periodíc Inspection ofCandu
r I I
Nuclear Power Components ASMEI
ASMEII ASME VIl
Boilers
ASME IV NBIC
This is an overwhelming list. And the task of sifting through the ASME V ASME IX
standards would have been monumental if it weren't for the appear-
ance in 2000 of API Recommended Practice 579 Fítness-for-Seruicell
r _l I
ASME VIII NBIC
which not only combines the best FFS methods from previous stan- ASME 11 API510
dards, but presents the process in a structured stepwise manner. Pressure Vessels
ASMEV API572
ASME IX API579
1.13 Cum Laude I 1 I

ASME 831.1 Ap.V
Once in a lifetime there appears a standard that is so timely and use- Power and ASME 831.1 API570
ful, so practical, so competent, and-what is rare-so well written that Process Piping ASME 831.3 AP1574
it represents a breakthrough in the profession. This is the case of API API579
Recommended Practice (RP) 579 Fítness-for-Seruice.n The document, I l 1
when first issued in 2000, contained eleven sections (1 to 11) and ten API128 API653
appendices (A to J). API650 API12R1
Storage Tanks ·• API620 STI SP001
1. Introduction AP12350 API579
2. Fitness-for-Service Engineering Assessment Procedure I 1 I

3. Assessment of Equipment for Brittle Fracture API610
Pumps API683
4. Assessment of General Metal Loss API674-676
5. Assessment of Local Metal Loss I 1 I

6. Assessment of Pitting Corrosion AP1570
ASME 816.34 API574
7. Assessment of Blisters and Laminations Valves
API600-609 API591
8. Assessment of Weld Misalignment and Shell Distortions API598
9. Assessment of Cracklike Flaws r -- I 1

.----- ASME I
10. Assessment of Components Operating in the Creep Regime Safety Relief ASME VIII
N81C
API2000
11. Assessment of Fire Damage Devices API576
AP1520-521
API526
A. Thickness, MAWP, and Membrane Stress Equations for an FFS I r - -· 1 I
Assessment 49CFR186-199 831.4
831.4 831.8
B. Stress Analysis Overview for an FFS Assessment Pipelines
831.8 831.85
API1104 49CFR186-199
C. Compendium of Stress Intensity Factor Solutions
D. Compendium of Reference Stress Solutions Figure 1.9 Hierarchy of codes and standards. 1-43
E. Residual Stresses in a Fitness-for-Service Evaluation 1.15 Response Ti me

F. Material Properties for an FFS Assessment The first challenge in equipment integrity is to establish the company-
G. Deterioration and Failure Modes level awareness and buy-in of the importance of Step 5 in Fig. 1.1, in
other words to inte1ligently scope high-risk systems, and take the time
H. Validation and spend the money to monitor their integrity. But once the first chal-
I. Glossary of Terms and Definitions lenge has been overcome, the second challenge appears very quickly:
J. Technical Inquiries to not let the inspection results sit on someone's desk, but instead to
have a quick response time.
Each degradation mechanism is the subject of a particular section. Not long ago, a storage tank leaked flammable materiais, causing a
Each section is structured in exactly the same manner, with three lev- fire, a fatality, and significant environmental damage to a nearby riv-
eis of assessment. er CFig. 1.10). Three weeks before the accident an "Unsafe Condition
Report" noted, "This Tank Farm Needs Attention Now!"
• Levei 1 is a conservative yet simple assessment, limited to simple The question at hand is: how long should it take to act on an inspec-
shapes (e.g., cylindrical shells) under simple loads (e.g., internai tion report? Response time should depend on (a) how critica] is the con-
pressure). sequence of failure, and (b) how significant is the degradation. This at
• Leve] 2 is a less conservative evaluation, involving more calcula- least is the approach taken by closely regulated inspection programs,
tions, and possibly more inspection accuracy, and applying to a such as those conducted in the pipeline and nuclear power industries.
broader range of shapes and loads (e.g., bending).
• Levei 3 is yet less conservative, but involves detailed analysis (often
finite element analysis), and possibly testing.
The chapters of API 579 also contain logic flow diagrams to help the
user successfully implement the procedure. The appendices are a com-
pilation of formulas, properties, and techniques necessary to apply
Chaps. 4 to 11.
All in all, API 579 Fitness-for-Service is an engineering gem; a docu-
ment that proves that clarity and competence go hand in hand. The
document does raise one caution: it is so welllaid-out and written that
it could be followed and implemented by a person with limited knowl-
edge in the five disciplines of fitness-for-service (mate1ials and corro-
sion, stress analysis and codes and standards, fabrication and welding,
inspection, and operation). And, obviously, that is a problem.
1.1 4 Technical Basis

Much of the origins and technical basis of today's codes and standards,
including fitness-for-service rules, can be obtained from an invaluable
series of technical reports published by the Pressure Vessel Research
Council (PVRC), in the form of "WRC Bulletins." These bulletins go
back to the 1940s, and are a comprehensive and practical source of
technical knowledge; they are meant to help the practicing engineer
understand the roots and the logic behind design and fitness-for-service
methods and margins. A list of WRC Bulletins is included in App. A. Figure 110 Tank fire caul'led by Jeak.~~
For example, in the pipeline industry, currently, the resolution of oil • Any other concerns judged to require 180-day action by the oper-
and gas pipeline inspection findings is related to the type and severi- ating company.
ty of the detected defect. s
In the nuclear power industry, periodic inspections are performed
1. Immediate action if during refueling outages, when the reactor is shut down. A safety sys-
• Over 80 percent of the pipeline wall is lost to corrosion. tem is not returned into service before the inspection results are ana-
• The calculated burst pressure at flaw is below the maximum lyzed and resolved.
operating pressure (MOP). Where there are no governing regulations, a good approach would be
• A dent is detected between 8 to 4 o'clock (top of pipeline), and to set, before the inspection takes place, response time leveis linked to
there is metal loss, a crack, or a stress riser. the severity of the inspection readings, for example:
• A dent is detected between 8 to 4 o'clock, and it is deeper than
• Green. Equipment OK as-is; evaluation would still be conducted to
6 percent of the diameter.
trend readings and determine next inspection interval;
• Any other concerns judged to require immediate action by the
operating company. • Yellow. Nonconfonnance report, to be resolved within 60 days;
2. Sixty-day action i f not an immediate condition, but • Red. Immediate action, possibly shut down system or equipment,
• A dent is detected between 8 to 4 o'clock, and it is deeper than 3 per- formal notifications, possibly additional immediate inspections to
cent of the diameter, or deeper than X in for 12 in and smaller confirm the problem and determine its extent.
pipelines.
• A dent is detected between 4 to 8 o'clock (bottom of pipeline), and 1.16 Summary
there is metal loss, a crack, or a stress riser.
• Any other concerns judged to require 60-day action by the oper- Fitness-for-service is a quantitative engineering method used to deter-
ating company. mine the remaining life of a system or component, and to make run-or-
repair decisions. Fitness-for-service, as described in this book, applies
3. One-hundred-eighty-day action if not immediate or sixty-day condi- to tanks. pressure vcssels, piping systems, pipelines, and more gener-
tions, but ally to static equipment. Fitness-for-service is but one step in the over-
• A dent is detected, and it is deeper than 2 percent of the diame- all process of compctent conduct of operations. Fitness-for-service
ter (or deeper than X in for 12 in and smaller pipelines), and it is applies primarily to existing degraded equipment, but it can also be
at a girth or seam weld. used to evaluate the severity of flaws and defects in new equipment, if
• A dent is detected between 8 to 4 o'clock, and it is deeper than 2 per- approved by all parties. API Recommended Practice 579 Fitness-for·
cent of the diameter, or deeper than Y. in for 12 in and smaller Service does an excellent job of laying down the step-by-step method
pipelines. for fitness-for-service assessment. This book describes the technical
• A dent is detected between 4 to 8 o'clock, and the dent deeper bases of fitness-for-service assessments, and illustrates its application
than 6 percent of the diameter. through practical examples.
• The calculated remaining strength of the pipe1ine reduces the
safety margin to less than 1.4.
References
• There is general corrosion deeper than 50 percent of the nominal
wall. 1. API Standard 1104, Welclíll!( of PipelineR and Related Facilities, American Petroleum
Institute, Washington, DC.
• There is wide circumferential metalloss. 2. Sperko, W.J., personal communication, Sperko Engineering Services Inc.,
• There is metalloss at crossing of another pipeline. Greensboro, NC.
• There is preferential metalloss at girth weld. 3. ASME Boiler and Pressurc VeRRel Code, Section VIII, Rules for Constructwn of
Pressure Vessels, Amcrican Socicty of Mcchanical Engineers, New York.
• A crack indication is confrrmed by excavation. 4. Code of Federal Regulations. 'Ntle 10, Energy. Part 50, Domestic Licensing of
• There is corrosion in a longitudinal seam weld. Production a11d Utliization flacilitíes, U.S. Govemment Printing Office,
• There is an indication of gouge or groove deeper than 12.5 percent Washington, DC.
5. ASME Boiler and Pressure Ves~el Code. Section XI, Rules for ln-Seruice lnspection
of the nominal wall. of Nuclear Pou·er Plants, Americnn Society of Mechanical Engineers, New York.
6. Code of Federal Regulations, Title 49, Transporlation, Part 192, Transportation of 34. ASME 8oiler and Pressure Vessel Code, Section li, Materials, American Society of
Natural Gas and Other Gas by Ptpeline: Minimum Federal Safety; Part 193, Mechanica\ Engíneers, New York.
Liquefíed Natural Gas Facilities: Federal Safety Standards; Part 194, Response 35. ASME 8oiler and Pressure Vessel Code, Section Ill, Rules for Construction of
Plans for Onshore Oil Pipelines; Part 195, Transportation of Hazardous Liquids Nuclear Facility Components, Ameiican Society of Mechanical Engineers, Ncw York.
Pipelines, U.S. Governmeni Printing Office, Washington, DC. 36. ASME 8oiler and Pressure Vessel Code, Section IV, Rules for Construction of
7. ASME 831.4, Pipeline Transportat10n Systems for Liquid Hydrocarbons and Other Heating Boilers, American Society of Mechanical Engineers, New York.
Liquids, American Society of Mechanical Engineers. New York. 37. ASME 8oiler and Pressure Vessel Code, Section V, Nondestru.ctive Examznation,
8. ASME 831.8, Gas Transmission and DistributLOn Piping, American Society of American Society of Mechanical Engineers. New York.
Mechanical Engineers, New York. 38. ASME 8oiler and Pressure Ves~el Code, Section VI, Recommended Rules for the Care
9. ASME 831.88, Managing System Integrity of Gas Pipelines. American Society of and Opcration of Heating Boilers, American Society of ~echanical Enginecrs, New
Mechanical Engineers, New York. York.
10. Code ofFederal Regulations, Title 29, Volume 5, Part 1910, Occupational Safety cmd 39. ASME 8oiler and Pressure Vessel Code, Section VII, Recommended Guid~lines for
Health Standards, Section 1910.119, Process Safety Marzagemenl o( Ilighly the Care of Power Boilers, American Society of Mechanical Engíneers, New York.
Hazardous Chemicals, Office of Safety and Health Administration COSHAJ, U.S. 40. ASME 8oiler and Pressure Vessel Code, Section IX, Welding and Brazing
Government Printing Office, Washington, DC. Qualification, American Society of Mechanical Engineers, New York.
11. API RP 579, Fitness-for-Service, American Petroleum lnstitute, Washington, DC. 41. ASME 8oiler and Pressure Vessel Code, Section X. Fiber-Reinforced Plasttc Pressure
12. API RP 12Rl. Setting, Maintenance, Inspection, Operation, and Repair ofTanks in Vessels. American Society of Mechanical Engineer::;, New York.
Production Service, American Petroleum Institute. Washington, DC. 42. NBIC, National Board lnspecti{)n Code, ANSIIN8-23, the National 8oard of Boiler
13. API 510, Pressure W!ssel Inspection Code: Maintenance, Inspection, Ratin!(, Repair, and Pressure Vessel Inspectors, Columbus, OH.
and Alteration, American Petroleum Institute, Washington, DC. 43. STl SPOOl, Standard for Inspection of In-Sert•ice Shop Fabricated Aboveground
14. API 520, Sizing, Selection and Installation of Pressure-relieving Deuices in Tanks for Storage of Combustible and Flammable Liquids, Steel Tnnk Tnstitute,
Refineries, American Petroleum Institute, Washington, DC. Lake Zurich, IL.
15. API RP 521, Cuide for l'ressure-relieving and D~pll!I:IIJLLring Sy:;lcm~:~, AIHcrkan 44. U.S. Chcmical Safety and Hazard Investigation 8oard, lnuestigation Report,
Petroleum lnstitute, Washington. DC. Refínerv Incident, Motiva Ente1prises LLC, Report No. 2001-05-1-DE. October, 2002.
16. API Standard 526, Flanged Steel Pressure Relzef Values, American Petroleum
Institute. Washington, DC.
17. API 570, Pipinl( Inspection Code: lnspection, Repair, Alterations, and Reratill(( ofln-
Sen•ice Piping Systems, Amerícan Petroleum Institute, Washington, DC.
18. API RP 572, lnspection o( Pressure Vessels, American Petroleum lnstitute,
Washington, DC.
19. API RP 574, lnspection of Piping, Thbing, Values, and Fittings, American Petroleum
20. API RP 591, Process Value Qualifícation Procedure, American Petroleum Institute,
Washington, DC.
21. API Standard 598, Value /nspection and Test. American Petroleum Institute,
Washington, DC.
22. API 600, Bolted Bonrzet Steel Gate Valt•es for Petroleum and Natural Gas Industrie.ç,
Modifíed Natwnal Adoption, American Petroleum Inshtute, Washington, DC.
23. API Standard 609, Butterfly Value.~: Double Flanged, Lu!(· and Wafer·Type, American
Petroleum Institute, Washington, DC.
24. API Standard 610, Centrifugai Pumps for Petroleum, Petrochemical and Natural
Gas, American Petroleum Institute, Washington, DC.
25. API Standard 620, Design and Construction of Large, Welded, Low Pressure Storage
Tanks, American Petroleum lnstitute, Washington, DC
26. API Standard 650, Welded Stcel Tanks for Oi/ Storage, American Petroleum
Instítute, Washington, DC
27. PI Standard 653, Tank lnspection, Repair, Alteration, and Reconstructwn Code,
American Petroleum Institute. Washington, DC.
28. API Standard 674, Positiue Displacement Pumps-Reciprocating, American
Petroleum Institute, Washington, DC.
29. API Standard 676, Positive Displacenumt Pumps· -Rotary, American Petroleum
30. API Standard 2000, Venting Atmospheric and Lou· Pressure Storage Tanks. Non·
Refrigerated and Refrigerated, American Petroleum Institute, Washington, DC.
31. ASME 831.1, Power Piping, American Society of Mechanical Engíneers, New York.
32. ASME 831.3, Process Piping, American Society ofMechanical Engineers, New York.
33. ASME 8oiler and Pressure Ves~el Code, Section I, Rules for Constmction of Power
Boilers, American Society of Mechanical Engíneers, New York.
Chapter
2
Materiais
2.1 Oemand and Capacity

Fitness-for-service relies on the comparison of the "demand" on the
degraded component ( the load exerted in-service in the form of pres-
sure, temperature, weight, vibration, flow transients, etc.) to the com-
ponent's "capacity" to sustain the demand. The capacity of the
degraded component depends on its mechanical properties. This chap-
ter focuses on understanding the key mechanical properties that
affect a component's fitness-for-service:
• Strength (yield stress and ultimate strength)
• Ductility Celongation at rupture or necking down of cross-section)
• Toughness (Charpy, fracture toughness)
2.2 Material Groups

Practically, materiais can be divided in two general categories: metal-
lic and nonmetallic. These can in turn be subdivided into groupings.
Standard categories and groupings, and their nomenclature, are
described in this chapter. More detailed descriptions of materiais and
their applications may be obtained from specialized publications and
textbooks.l-4
Figure 2.1 is a simple diagram o f metallic materiais commonly used
for tanks, vessels, pipe, tubing, or pipelines.
25
26 ChapterTwo Materiais 27
• Gray cast iron is a cast iron produced by slow cooling of the iron from
Metallic the melt, with a large proportion of graphite in the forro of flakes in
a matrix offerrite and pearlite. Gray iron castings for valves, flanges,
and pipe fittings have an uitimate strength ranging from 20 to 40
I ksi, but tend to be brittle because they have no guaranteed yield or
Ferrous Nonferrous elongation at rupture. 7
• Ductile iron is a rapidly cooled cast iron with a large proportion of its
graphjte in spherical nodules. Ductile iron castings have good
r strength, with an ultimate strength around 60 ksi, similar to grade
Cast lron Steel Nickel B carbon steel, and good ductility, with a guaranteed minimum yield
stress of around 40 ksi, and 15 to 20 percent elongation at rupture.
• White iron is cast iron with carbon in the combined forro of cemen-
I f-- Copper tite Fe3C. Malleabie iron is a white cast iron that has been
Carbon ,.....-- Alloy
annealed, with a large proportion of its graphite in elongated clus-
ters that are ductile but still maintain a good hardness.
f-- Titanium Steel is a metallic material with more iron than any other element,
f-- Low Alloy and Iess than 2% carbon, otherwise it wouid be Iabeled cast iron.
Carbon steel is steel with a prescribed maximum weight content of 2%
'-- Aluminum carbon and 1.65% manganese. Carbon steel, bare or painted, is the
I--
lntermediate workhorse ofthe process industry (Fig. 2.3). There are several types of
Alloy carbon steeis:
• Low-carbon steeis, aiso referred to as mild steeis, have a carbon con-
'-- HighAIIoy tent below 0.3%; they are commonly used in making tanks, vessels
and pipe.
Figure 2.1 Common metallic materials diagram. • Medium-carbon steels have a carbon content between 0.3 and 0.6%;
they are commonly used in railway applications.
2.3 Ferrous Metais
Ferrous materiais are iron-based metais. They consist oftwo large cat-
egories:5
• Cast irons: white, gray, ductile.

• Steeis: Iow-, medium-, and high-carbon, alloy steels.
The term cast iron describes a series of iron and carbon alloys with a
carbon content in excess of 2%, weight (in describing chemical composi-
tion of metais, throughout, percent refers to the percentage of an element
in weight). Cast irons usually contain close to 1% silicon to improve
machining. Cast iron is commonly used for the body of valves, for pump
casings, and for underground waterworks (Fig. 2.2). There are different
groups of cast irons, with varying microstructures and properties:6 Figure 2.2 Brittle fracture of a cast iron water pipe.
TABLE 2.1 Common ASTM or ASME 11 Boller Materials67
Material Tube Pipe Casting Forging Piate
Carbon steel 178, 192, 210, 53, 106 216 105 299.515
226
C-0.5Mo 209
1Cr-0.5Mo 213-Tl2 335-Pl2 182-Fl2, 387-12Cl2
336-Fl2
1.25Cr-0.5Mo 213-Tll 335-Pll 217-WCG 182-Fll 387-11C12
2.25Cr-1Mo 213-T22 335-P22 217-WC9 182-F22 387-22Cll,
387-C12
5Cr-0.5Mo 213-T5
9Cr-1Mo 213-T9
18Cr-8Ni 213-TP304H 376-TP304H 182-F304H 240-304,
240-304H
Figure 2.3 Painted carbon steel-the workborse of plant materiais. 213-321H
18Cr-10Ni-Ti
18Cr-10Ni-Cb 213-347H
• High-carbon steels have a carbon content between 0.6 and 1.0%; 16Cr-12Ni-2Mo 213-TP316H 376-TP316H 182-F316H 240-316H
they are commonly used in making steel wires. 25Cr-12Ni 351-CH20
Alloy steel is steel, other than stainless steel, with a mtmmum

weight content of elements such as 0.3% aluminum, 0.3% chromium,
0.4<k copper, 1.65% manganese, 0.08% molybdenum, 0.3Cif nickel, 0.6% High-alloy steels, such as stainless steel, have a chromium content
silicon, 0.05% titanium; 0.3% tungsten; 0.1% vanadium; except for sul- in excess of 10%, and carbon content below 1.20%.
fur, phosphorus, carbon, and nitrogen. At room temperature, alloy steels with chromium content in excess of
Through practice and experience, industries have developed pre- 10.5% form a highly adherent, thin chromium oxide layer which passi-
ferred alloy applications, such as listed in Table 2.1 for boilers. More vates the steel, and renders the metal "stainless." Stainless steels are
general steel material specifications are listed in Table 2.2. alloy steels with chromium in excess of 11%, with nickel and lesser quan-
tities of other elements. They can be ferritic, martensitic, austenitic or
• Low-alloy steels as steels with less than 5% total alloys. Examples duplex (ferritic-austenitic). Stainless steel is an essential alloy in corro-
of low-alloy steels include 0.5Cr-0.5Mo (ASTM A 335 P2), 1Cr- sion-resistant applications. It is readily formed, machined and welded,
0.5Mo (ASTM A 335 P12), 1.5Cr-0.5Mo (ASTM A 335 PU), 2Cr-1Mo but it can be sensitized by precipitation of chromium carbides at grain
CASTM A 335 P3b), 2.25Cr-1Mo (ASTM A 335 P22), 3Cr-1Mo boundaries (Chap. 5). This precipitation process depletes chromium from
(ASTM A 335 P2l).s& the grain boundaries and therefore renders the metal susceptible to
• Intermediate-alloy steels contain between 3% and 10% Cr, such as intergranular attack in these depleted regions. Stainless steel is com-
4 to 9Cr-0.5 to 1Mo (ASTM A 335 P5 to P9). monly used for corrosion resisting or high-temperature service tanks,
• High-strength low-alloy steels are carbon steels with microalloys vessels and piping (Fig. 2.4).
and fabrication processes that give them high yield and a tensile Stainless steels come in different forros, each suitable for particular
strength of 100 ksi or more, they are commonly used in shipbuilding. applications:
• Common structural materiais are A36 for carbon steel, and A167 for • Austenitic stainless steels (300 series). They can be procured with low
stainless steel. carbon (with the L suffix, such as 304L or 316L, with 0.03% or less
30 Chapter Two Materiais 31
TABLE 2.2 Common ASTM Steel Groups
Material Pipe or tube Casting Forging Plate

Carbon steel 53, 106, 120, 216,352 105, 181, 234, 283,285,299,
134, 135, 139, 268, 350, 372, 442,455,515,
178, 179, 192, 420,508,541 516,537,570
210, 211, 214,
226, 333, 334,
369, 524, 587,
671,672,691
C-0.5Mo 209, 250, 335, 217,352,487 182, 234, 336, 204,302,517,
369, 426, 672, 508,541 533
691
1Cr-0.5Mo 213, 335, 369, 184,234,336 387,517
426,691
1.25Cr-0.5Mo 199, 213, 335, 217 182, 234, 336, 387,517
369,426,691 541
2Cr-0.5Mo 199,213,369
2.25Cr-1Mo 199, 213, 335, 217,487 182, 234, 336, 387,542
369,426,691 541,542
3Cr-1Mo 199, 213, 335, 182,336 387
369,426,691
5Cr-0.5Mo 199, 213, 335, 217 182,234,336 387
369,426,691
7Cr-0.5Mo 199, 213, 335, 182, 234 387 Figure 2.4 Stainless steel process
369,426 columns.
9Cr-1Mo 199, 213, 335, 217 182,234,336 387
369,426
Stainless 213, 249, 268, 351 182,336,403 167,240,412,
269, 312, 358, 457
376, 409, 430, • Duplex stainless steels (alloy 2205, 3RE60). They are high chromium-
451,452 low-nickel steels with a dual microstructure: ferritic and austenitic,
and are meant to have the advantages of both: high wear resistance
and good corrosion resistance. Their use has greatly increased in the
carbon) to mmnruze the risk of grain boundary precipitation of chemical and petrochemical industries.
chromium carbides during welding, or high carbon (with the H suf-
• Precipitation-hardened stainless steels (martensitic 17-4PH, semi-
fix, with 0.04% or more carbon) for higher strength, particularly at
austenitic 15-7PH, AM350, austenitic 17-7PH). Unlike austenitic
high temperature. Austenitic stainless steels have relatively high
stainless steels, the "PH" types can be hardened by heat treatment.
ductility and toughness. They are commonly used in corrosive and
Initially used in aerospace applications, they are becoming more
high-temperature utility and process servi.ces.
common in process plants.
• Martensitic stainless steels (410, 416). These are straight chromium
steels, with no nickel. They have good hardness and wear resistance.
2.4 Nonferrous Metais
• Ferritic stainless steels (405, 409, 4108, 430, 446). Like martensitic
steels they are straight chromium steels. They are generally resis- Nonferrous metais are non-iron-based metais, such as aluminum and
tant to stress corrosion cracking and are commonly used in automo- its alloys, nickel and its alloys, copper and its alloys, and titanium and
tive applications. its alloys.
Nickel alloys. Nickel is a metal with good corrosion resistance, and

good strength at high temperature. It is readily welded but, as was
the case with stainless steel, it can be sensitized. Nickel is most
often used as an alloy in stainless steel, but is commonly used in the
form of nickel alloy, for high-temperature processes and furnaces.
Much useful information on nickel and its applications may be
obtained from the Nickel Institute, Toronto. 4 Nickel alloys include
the following.
• Pure and low alloys: 200, 301.
• Ni-Cr-Fe alloys: 600 and 800 series. Nickel-chromium alloys (such
as Inconel® or Hastelloy®) are selected for high-temperature ser-
vice, up to around 2100°F, where they maintain good corrosion
resistance and good strength compared to steel.
• Ni-Cu alloys: 400 series. Nickel-copper alloys such as Monel® are
well suited for reducing environments and seawater.
Copper alloys. Copper is a soft metal, easily bent and welded, Figure 2.5 Titanium heat exchanger tubes.
10
brazed, or soldered. lt is generally resistant to corrosion in potable

or salted water and caustic solutions. Copper may be alloyed with
zinc (brass), with tin (tin-brass), or with lead (lead-brass). Bronze treated or work-hardened aluminum alloys lose their mechanical
is a copper-tin alloy with the addition of phosphorous, aluminum, or properties when welded.
silicon. Copper-nickel alloys (cupronickel) are used in vessels and
heat exchangers.57- 65 2.5 Nonmetallic Materiais
Titanium. Titanium is a lightweight, corrosion-resistant metal, Nonmetallic materiais used in tanks, vessels, or p1pmg systems
forming a stable passive oxide film Ti02 , resistant to many acids, include plastics, concrete, glass, and glass- or graphite-reinforced
sulfides, and chlorides. It also has good resistance to pitting, crevice materiais such as fiber-reinforced plastics (FRP), and ceramics.
corrosion, and microbial corrosion. It can be used pure or alloyed Plastics can be thermoplastics or thermosetting resins. Thermo-
with aluminum, vanadium, and molybdenum. Titanium and alloys plastics are materiais that can be repeatedly softened when heated,
are common in aeronautical and aerospace applications and in cor- without effect on their properties. They include PVC (polyvinyl chloride),
rosive chemical processes (Fig. 2.5). polyethylene (PE), high-density polyethylene (HDPE), chlorinated PVC
Aluminum. Aluminum is a corrosion-resistant material with a high (CPVC), acrylonytrile butadiene styrene (ABS), styrene rubber (SR),
strength-to-weight ratio that can be further strengthened through polybutylene (PB), polypropylene (PP), polyvinyldiene chloride (PVDC,
heat treatment or cold working. Aluminum alloys are identified by a Saran®), fluoroplastics such as polyvinyld.ine fluoride (PVDF) or polyte-
four-digit number. The first digit refers to the alloy group l:XXX: for trafluorethylene (PTFE, such as Teflon® or Halon®) or ethylene chlorotri-
pure (over 99%) Al, 2:XXX: Cu alloy, 3:XXX: Mn alloy, 4:XXX: Si alloy, fluoroethylene (Ralar®), styrene-rubber (SR), chlorinated polyether
5:XXX: Mg alloy, 6:XXX: Mg + Si alloy, 7XXX Zn alloy, and 8:XXX: oth- (CPE), cellulose acetate butyrate (CAB), and polycarbonate (Lexan®).
er alloys. The temper designation is indicated by a letter, such as F Thermosetting plastics are materiais that set when cooled (they are
as-fabricated, O annealed, H strain hardened, W solution heat- said to "cure") and cannot be repeatedly softened by heating without
treated, and T heat-treated. For example, the temper designation T6 altering their properties. They include epoxies, phenolics, and poly-
applies to aluminum that is solution heat-treated then artificially esters. Thermosetting plastics are often used as coating, lining, or in
aged. Aluminum maintains good strength and toughness down into combination with powders or fibers to form fiber-reinforced plastics.
the cryogenic range (-450°F liquid helium). Resistance to stress cor- Resins commonly used in FRP piping include epoxy, polyester,
rosion cracking varies based on alloy and heat treatment.8•9 Heat- vinylester, and phenol-, urea-, melamine-formaldehydes. 11-14
34 Chapter Two
Materiais 35
2.6 Basis for Material Selections

toughness is an option that may be called for by the owner or the
Materiais for tanks, vessels, piping, and pipelines are selected on the designer, in the form of a Charpy V-notch test or a drop-weight tear
basis of four primary characteristics: test; for example, supplementary requirements may be specified as
follows:
• Corrosion resistance: resistance to wall thinning, pitting, cracking,
and metallurgical transformation (Chap. 5) • "Charpy V-notch impact tests shall be conducted in accordance with
• ~hop. and ~eld fabricability: ease of transport, assembly, and erec- Specification ASTM A 673. The frequency of testing, the test tem-
tiOn, mcludmg bending, machining, welding, and coating (Chap. 4) perature to be used, and the absorbed energy requirements shall be
• Mech~cal properties: strength, ductility, and toughness, through the as specified on the order."
operatmg temperature range (from creep down to cryogenic service) • "Drop-weight tests shall be rnade in accordance with Test Method
• Cost ASTM E 208. The specimen s shall represent the material in the
fmal condition of heat treatment. Agreement shall be reached
between the purchaser and the manufacturer or processar as to the
2.7 Mechanical Properties Overview number of pieces to be tested and whether a maximum nil-ductility
Materiais have two mechanical properties of interest to fitness-for- transition (NDT) temperature is mandatory or if the test results are
service: strength and tonghness. for information only."
• ~tren~h is ~he material's yield stress Sy (elastic-to-plastic transi-

tton), Its ulttmate strength Su (rupture stress in tension) and its 2.8 How to Achieve Desired Properties
elongation at rupture eu (it can be argued that elongation at rupture The desired mechanical properties, strength and toughness, can be
s?oul? be r~ferred to as ductility). The minimum strength of mate- achieved in severa] ways:
nals 1s an mtegral part of material specifications. When we order
materia~s to ~ certain specification and grade we automatically get • Alloying, also referred to as solid solution strengthening. When
a matenal Wlth str ength equal to or larger than a minimum value added to steel, alloymg elements such as carbon, sili.con, and man-
de~ned in the material specification. For example, material specifi- ganese reduce the slippage of dislocations in the metallic crystal;
catwn ASTM A 573, grade 70 is a common carbon steel plate for stor- this in turn increases the material's strength (yield stress and ulti-
age ~a.nks; th~ material specification requires, at room temperature, mate strength). The str ength of steel can also be improved by the
a mmun~ Yleld str~s~ Sy = 42 ksi, a minimum ultimate strength addition of small quantities (microquan tities) of specific alloys. The
Su = 70 ks1, and a nummum elongation at rupture eu = 21 percent most common microalloying elements are niobium (0.10%), vanadi-
in a 2-in-long specimen. um (0.10%), and aluminum (0.03%). But, as is often the case in met-
• Toughness is the ability of the material to resist impact and to hold allurgy, improvements in strength are often achieved at the cost of
~ crack stable under load, and not to let the crack rip open. To help lower toughness and lower weldabili.ty.
1llustrate otoughness, we compare a piece of steel at room tempera- • Dislocation strengthening, also referred to as work hardening.
ture of 70 F and the same steel in a very cold environment at -70°F. Mechanical properties can vary by cold working. The plastic defor-
It is i~tuitivel?' evident that the cold metal ruptures m~re easily mation of metais below the lower criticai temperature often results
under 1mpact; 1t has a lower toughness. Toughness is measured in a in an increase in hardness and strength . For example, a pipeline
variety of ways, the most common being the Charpy V-notch impact that is bent cold in the field undergoes strain hardening in the cold
test, the fracture toughness test, and the drop-weight tear test bend. Another example is the cold expansion of line pipe to its final
described later in this chapter. ' size: th e expansion process causes dislocations to be locked in place,
increasing the material's strength.
In a few material specifications, toughness is a standard material
requirement, but in most specifications toughness is either optional or • Optimizing the microstructure and grain refinement, also refeiTed to
it is not mentioned. For example, for an ASTM A 573 tank steel plate, as precipitation strengthening. A more refined grain size generally
leads to a more ductile material. Grain size is measured following
standard practices, and assigned a number: the larger the number, lron Carbon Equllibnum Phase 0Wlgram
'"' 0 .3% Cart>on Steet
the finer the grain. One definition of the grain size number G is56
N = za-1
where N = number of grains per square inch at lOOX magnification
G = ASTM grain size, from 0.0 (500 ll. diameter grain) to 14.0
(2.8 fJ. diameter grain).
2.9 Phase Diagram of Carbon Steel

To understand how the grain size, the microstructure, is affected by
fabrication of the base metal or by welding, consider the equilibrium
phase diagram of carbon steel (Fig. 2.6).
Starting with a carbon steel plate with 0.3% carbon at room tem-
perature, we place a point on the horizontal axis at 0.3, in Fig. 2.6.
This point represents the steel plate at room temperature. When the
plate is heated in a furnace or welded, the point follows the vertical
line upward at 0.3% carbon, until the point enters the "liquid alloy"
zone, around 2800°F, the melting temperature of steel. At this point
the metal in the weld bead is a mo]ten liquid pool of atoms. As the
metal or weld cools down, the molten pool solidifies (white-hot steel),
keeps cooling down, and enters the austenitic solid solution zone,
around 1700°F (the metal is now orange-hot). In the austenitic zone,
the atomic structure of the metal grain is face-centered-cubic (fcc;
Fig. 2. 7), with iron atoms at the corners of a cubic lattice and one
atom in the middle of each cube face. As the metal continues to cool,
the point keeps moving down the heavy verticalline until it exits the Colora & 0.3 0.5
Grafn Size
austenitic region, around 1400°F (the metal is red-hot); this is a zone , . _ _ HypoEutectOid HyperEutectood - + .______ HypoEutecUc HyperEutect>c
where the grain is smallest (most refined), also, at this point, the fcc -----.j ~~4---- Stoe4s ----~~~ Semi-Steels 14 Castlrons - -- - .
1rons
austenite converts to body-centered-cubic (bcc) ferrite and pearlite % Carbon
(lamellar structure of ferrite and iron carbide; Fig. 2.8) and stays Figure 2.6 Phase diagram of steel. 4
ó
that way until the metal is fmally back down to room temperature.
This sequence of events occurs if cool-down is slow; in practice this
is what happens when thin sections (less than approximately half an
inch) cool in ai r, or thicker sections cool in a furnace. But if the cooling
rate is more rapid, such would be the case when white-hot austenitic
earbon steel is quenched in water; the fcc atoms in the austenite have
no time to fully rearrange themselves into the bcc ferritic structure of
Fig. 2.8, and the end product is a distorted microstructure, with high
residual stresses and high hardness, called martensite (Fig. 2.9). This
hard martensitic structure lacks ductility, and tends to crack by
shrinkage or by absorption ofhydrogen (Fig. 2.10). Figure 2.7 Face-centered cubic
In light of the phase diagram of steel, we can understand the austenite.
microstructure of a single-pass carbon steel weld, in the as-welded
5 Weld pool
3000 F - - - -
4
- - - - 2000"F
- _:}_- 1650"F
2
11 OO"F
Figure 2.8 Body-cent.ered cubic

ferrite.
Figure 2.11Simplified view of

weld microstructure.
condition. The microstructure, depicted in Fig. 2.11, will comprise the

following regions, where zones 2 to 4 are the "heat-affected zones"
(HAZ):
• Zone 1 is sufficiently rcmoved from the weld; its microstructure

remains that of the base metal.
• Zone 2 is referred to as the intercritical region, which has only expe-
rienced a partia! transformation to austenite.
• Zone 3 is a fine grain region; the metal has been above the lower
Figure 2.9 Atomic structure of criticai temperature in the lower end of the austenitic region.
martensit.e
• Zone 4 is a coarse grain region; the metal has been in the high-
temperature end of the austenite region, forming large grains.
• Zone 5 is the weld pool, a casting for ali practical purposes.
2.1 O Heat Treatment

The formation of excessive martensite during welding offerritic steels
and subsequent cracking can be prevented by pre- and postweld heat
treatment:
• Preheating the metal before welding, causes the weld, surrounded

by preheated hot or warm metal, to cool down more slowly (Fig.
2.12).
• Preheating the metal before welding wil1 drive off hydrogen con-
taining moisture, and welding with low-hydrogen electrodes will
Figure 2.10 Hydrogen-induced crack.ing of weld.15 prevent hydrogen diffusion into the weld.
co
I
1/)
1/)
Q)
c
'E
tO
.s:;
~200
c
;i5
Weld
Figure 2.13 Effect ofheat treatment on hardness (see Table 2.3).
Figure 2.12 Heating blankets for preheat and interpass

heating.
TABLE 2.3 Effect of Pre- and Postweld HeatTreat on Hardness 1Y.Cr-l4Mo
Maximum hardness <Brinell> Preweld beat, °F Postweld heat, °F

• Postweld heat treating below 1400°F will relieve residual stresses.
300 Nane Nane
• Postweld heat treating above 1400°F will transform the martensite
275 400 Nane
back to austenite and then controlled cooling will prevent the
mattensite from reappearing. 225 Nane 1300
200 400 1300
Hardness measurements can then be used to confirm that the heat
treatment has achieved its objective of transforming the martensite
back to ferrite, as illustrated in Fig. 2.13 and Table 2.3. The figure
compares the effect of four heat treatments on the hardness of the • Grain refinement and tempering of the weld heat-affected zone,
heat-affected zone of a weld between two pieces of l X Cr-Y2 Mo steel. which improves strength and toughness
A word of caution regarding hardness of the weld heat-affected zone: • Reduction of weld and heat-affected zone hardness to original base
The surface of the heat-affected zone may be decarburized and there- plate hardness, to prevent cracking
fore provide hardness readings that are deceptively low compared to
the actual material underneath the thin decarburized outer layer. Heat treatment should be limited to the desired phase transforma-
tion temperature, around 1400°F for carbon steel, and not much more,
in order to achieve a refined grain size. If the heat-treatment temper-
2.11 Benefits of Postweld Heat Treatment ature is too high in the austenitic zone the grain will grow excessively.
Postweld heat treatment is imposed in construction codes, often as a
function of material type and thickness. If performed correctly, heat 2.12 Types of Heat Treatment
treatment has several benefits:
Heat treatment, in general, is referred to as annealing.5 Common heat
• Reduction of residual stresses, which reduces distortions and risk of treatments include: normalizing, quenching, stress relieving, and tem-
cracking pering.
around the weld. The beads may be placed in stainless steel boxes
(bands) with one open side facing the weld, or they may be suspend-
ed from the vessel wall with tie-wire.
---- N • Heating with diesel fuel burners placed inside large vessels (Fig.
u.
0 I ',
~ 1500 -------------L-------~-------- 2.16).
___________ J ___________
1 ' 1
~ 1333 • Heating with a handheld torch. The heating temperature is difficult
.\
to control under these conditions, which is why torch heating is usu-
~ 1000 .1I \ \ ally limited to preweld heating to dry the weld area and heat the
Q) SR \
f- metal to prevent thermal shock or rapid cooling when welding in a
'o ' \
I cold environment.
\
\
200 Cautions when heat treating are as follows:
' \
6 8 13 • Attach thermocouples to the component to monitor temperature ver-

Time, h sus time. Thermocouple readings should be constantly monitored
and correlated to the section and heater number (Fig. 2.17).
Figure 2.14 Postweld heat treatmcntR.
• Account for the expansion and contraction of equipment.
• In vertical vessels, prevent the heat-treated sections from buckling
• Nomalizing (N in Fig. 2.14) is a heat treatment in which steel is under the weight of metal above it.
heated abov~ the ferrite- ·austenite transformation range (approxi-
mately 1470 F), and then cooled in air to a temperature substan- • Use insulation around heat-treated areas, with ceramic-wool and
tially below the transformation range. mineral wool or fiber (Fig. 2.18).
• Quenching (Q in Fig. 2.14) is fast cooling. For example, during fab- • PT, MT, UT, or RT the heat-treated area, and adjacent welds, for evi-
rication of stainless steel flanges, the white-hot forging is quenched dence of cracks after PWHT.
by being pushed from the heating oven directly into a water tank. • Wear heat-protective suits, and wear masks ar ound insulation.
• Stress relieving (SR in Fig. 2.14) is heating a metal, holding it hot
long enough, and then cooling slowly, with the purpose of reducing
built-in stresses.
• Tempering is reheating a quench-hardened or normalized steel
object to a temperature below the lower criticai temperature, and
then cooling it at a controlled rate.
2.13 Shop and Field HeatTreat ment

Heat treatment of components and subassemblies can be performed in
severa) ways:
• Heating in a fumace in the shop or in the field CFig. 2.15). When heat
treating in the field, an open bottom oven is assembled, lifted. and
lowered around equipment.
• Heating with electric resistance blankets or nickel-<:hrome wire
heaters placed inside ceramic beads. The beads can be wrapped Figure 2.15 Shop heat-tl·eatment oven.
44 Chapter TWo Materiais 45
@TC
8 G
Figure 2.16 Field heat treatment of vessel with burners <TC:

thermocouples, B· burner control, G: gas).l6
Figure 2.18 Postweld heat trcatment.68
2.14 The larson-Miller Parameter

The benefits of postweld heat treatment are a function of temperature
and time at temperature. Figure 2.19 compares the ductility (elonga-
tion at failure) for 1.25Cr-0.5Mo steel, under three weld heat-treated
conditions: as-welded at top, heated at 1100°F for 1 h (middle), and
heated at 1300°F for 1 h (bottom).
Metallurgists often measure heat treatment in terms of a parameter
that captures heat-treatment tcmperature and time at temperature,
the Larson-Miller parameter LMP, defined as
LMP = (T 460)(20 + log t) X 10-3
where T = temperature, °F, and t = time, h.
2.15 Heat and Lot

A heat is to metallurgy what dough mix is to bak.in g, and a lot is the
final batch of baked goods:
• A heat is a quantity of steel, before it is turned into product fo1·ms. Ali
Figure 2.17 Field-heating arrangement.ss parts made from the same heat \vill have similar chemistry (similar
Iots of flanges and caps. In this case, the same heat of steel was used
to make different product lots; each lot has the same chemistry, but
each lot has its own fabrication and heat-treatment sequence, and
therefore its own microstructure and mechanical properties.
2.16 The Three Strength Parameters

Strength is the relationship between strain (elongation per unit
length) and stress {applied load per unit area). This relationship is
obtained through a tensile test in which a specimen of standard size
(Fig. 2.20) is placed in a tensile machine and pulled.l7,18 The specimen
will stretch, first elastically (linear stress-strain portion of the curve,
Fig. 2.21) until it reaches a yield point (the yield stress Sy, around
38 ksi in Fig. 2.21). In the elastic region, the relationship between
stress and strain is linear, and can be written simply as
a= Ee
where cr = engineering stress, psi

E = modulus of elasticity (Young's modulus), psi
e = engineering strain
F
<J=-
Ao
where F = applied force, lb
A0 = initial cross section of tensile specimen, in2
!1L = specimen elongation, in
Lo = initiallength of specimen, in
Figure 2.19 Increased ductility with time and temperature.ts In this classic and practical description of stress and strain, force divid-
ed by initial area and elongation divided by initiallength, the stress and
strain are called "engineering'' stress and strain.
ingredients), but not necessarily the same microstructure or mechan- Beyond the yield stress Sy, the material will deform plastically (the
ical properties. strain deformation of the specimen is no longer proportional to the
• Unlike chemistry, impurities or defects such as porosities are not applied stress), all the while the specimen is necking down at its cen-
homogeneous within a heat; they segregate and concentrate in par- ter. Upon reaching a maximum stress (the ultimate strength Su,
ticular areas during the formation of an ingot, casting, or plate. around 63 ksi in Fig. 2.21) the specimen has lost its load-carrying
capability, and will rupture shortly afterwards. The final elongation
• A lot is a quantity of a product (tubes, pipes, flanges, fittings, plates, of the specimen when it ruptures is the elongation at rupture eu,
etc.) manufactured under uniform conditions, and therefore a lot sometimes referred to as ductility (a little over 0.20 or 20 percent in
tends to have a similar microstructure and similar mechanical prop- Fig. 2.21).
erties.
It is diffi.cult to describe the relationship between engineering stress
For example, a pipe rnill buys from the steel mill a heat of metal to and engineering strain in the plastic region. To circumvent this prob-
a certain specification. With that heat the pipe mill may make several lem, we define the "true" stress and true strain as
Et = true strain
!iL(t) = elongation ata moment in time t, in
L(t) = specimen length at a moment in time t, in
In the plastic region, the relationship between true stress and true
strain is
where k = strength coefficient, psi, and n = strain hardening coefficient.

This relationship can also be written in terms of parameters o. and
n called the Ramberg-Osgood parameters
Figure 2.20 Slainless steel tensile test specimen before and after test.
where e = :sl.raiu
I
60 -- -----I- Eo = reference strain = crofE
E = Young's modulus, psi
50 cr = stress, psi
cr0 = 0.2 percent offset yield stress, psi
'iii I
~ 40 -~-------
I I
- - - - - -.- - - - - - ... - j ' - - -
In summary, the material has a yield stress Sy (expressed in ksi) ,
ui
!/)
~
an ultimate strength Su (expressed in ksi), and an elongation at rup-
üS 30 ture, ductility, eu (expressed in percent). These three mechanical prop-
I I
erties constitute the strength of the material. Minimum values of Sy,
20 ---~-------~- ------- ~-------·--- Su, and eu are imposed on the material manufacturer through mater-
1 I
ial specifications, such as ASTM or API specifications.
10
2.17 Allowable Stress
o
2.5 5 10 15 20 The ASME design rules for tanks, vessels, piping, and pipelines rely
Strain,% on maintaining a safety margin between the maximum calculated
Figure 2.21 l<~xample of stress-strain curve for mild carbon steel. stresses in the component and a limit based on the material yield
stress Sy or ultimate strength Su at the corresponding temperature.
This limit is referred to as the code allowable stress, and is noted S or
Sm depending on the code. For pipelines, the ASME B31.4 and ASME
F(t) liL(t) B31.8 allowable stress is simply 72 percent Sy. For process plant pip-
(J =-- E=--
1 A(t) t L(t) ing, the ASME B31.3 allowable stress is the lowest of the following
values:
where cr1 = true stress, psi
F(t) tensile force applied to specimen ata moment in time t, lb • One-third of Su at temperature
A(t) = cross section area of tensile specimen at a moment in • The lower of two-thirds of Sy at room temperature and two-thirds of
time t, in 2 Sy at temperature
TABLE 2.4 Allowable Stress 831.1 TABLE 2.5 Approximate Correlatlon Hardness, Ultlmate Strength
400°F 600°F 800°F 1000°F Brinell Vickers Rockwell B Rockwell C Su, ksi
A 106 Gr. B 15 15 10.8 269 284 27.6 131
A335 Gr. PU 15 15 14.4 6.3 229 241 98.2 20.5 111
A 376 Type 316 13.4 11.8 11.0 10.6 187 196 90.9 90
B 167 Ni-Cr-Fe 17.2 16.3 15.5 14.7 143 150 78.6 71
121 127 69.8 60
• For austenitic stainless steels and nickel alloys having similar

stress-strain behavior, the lower of two-thirds Sy at room tempera-
ture and 90 percent of Sy at temperature
• 100 percent ofthe average stress for a creep rate of0.01 percent per
1000 h
• 67 percent of the average stress for rupture at the end of 100,000 h
• 80 percent ofthe minimum stress for rupture at the end of 100,000 h
Table 2.4 compares the strength of seamless pipe as a function of
temperature, in this case on the basis of ASME B31.1 allowable stress.
Whereas up to 600°F the strengths are comparable, notice the pro-
gressive improvement of high-temperature strength, above 600°F,
when going from plain carbon steel, to low-alloy steel, stainless steel,
and nickel alloys.
2.18 Obtaining Strength Properties

of Operating Equipment
For operating equipment, strength properties (Sy, Su, and eu) can be
obtained from the original material specification or the original mate-
rial certificates of the component, when these are still available. If
these two sources are not available, then testing is necessary. The sim-
plest test is a hardness test of the surface, and the estimation of an Figure 2.22 Removed boat sample.68
approximate ultimate strength Su from the hardness readings, as
shown in Table 2.5.
An accurate measurement of Sy, Su, eu and the full stress-strain
curve can be obtained by taking a "boat sample" from the actual com-
ponent (Figs. 2.22 and 2.23) and placing it in a tensile test machine.
This presumes that the metal has excess thickness that permits
removing the sample, or that the boat sample cutout can be replaced
by weld deposition.
Strength properties can also be measured in place by nondestructive
techniques. In particular, stress-strain microprobes (SSM) utilizing an
automated ball indentation machine (ABI) can be used to obtain, in Figure 2.23 Boat sample used for tensile testing.19
situ, mechanical properties such as yield strength, flow stress, strain-

hardening exponent (plastic stress-strain), and fracture toughness. 20•21
2.19 Factors Aftecting Strength Properties

The strength properties of the material depend on four factors:
• The chemical composition and microstructure, which are set by the
material specification, mill and shop fabrication, and field erection.
• The fabrication and construction ofbends and welded joints.
• The strain rate of the applied load. Stress and strain r ates are stan- Figure 2.24 Necking and cup-
dardized through tensile test procedures such as ASTM E 8. 18 and-cone fracture surface.
Standard tests are conducted at a stress rate of 10,000 to 100,000
psi!min to yield, anda strain rate between 0.05 and 0.5 1/min from
yield to ultimate. For steel, strength increases with strain rate; at
very high strain r ates, on the order of 100 s-1, yield stress Sy can The important point is to communicate how ductility is defined.
nearly <.luul>le and ultimate strength Su can increase by nearly 50 Ductility can be defined as elongation at rupture, or necking of t~e .cross-
percent.22 sectional area of a tensile specimen (Fig. 2.24). A rupture stram m the
order of eu - 15 percent is for ali practical purposes a proof of ductility.
• Temperature, which is set by the operating conditions. The relation- Such is the case for the stainless steel specimen of Fig. 2.20, where the
ship among yield stress, ultimate strength, and temperature for many ruptured specimen appears to have stretched close to 20 percent com-
metais is listed in standards such as ASME Boiler and Pressure
pared to its initiallength.
Vessel Code, Section II, Materiais. Table 2.6 provides examples of Notice in Fig. 2.24 how the ductile fracture is characterized by neck-
variability of yield stress with temperature ing-down, and a "cup-and-cone" shape of the fractured surface.
2.20 Ductility
2.21 Ductile Fracture
There is no universal definition of ductility. Some engineers define A material with high yield stress Sy will be able to deform elastically
ductility on the basis of toughness or shear area at fracture; others under high load, and, if it remains elastic (applied stress below the
define ductility on the basis of elongation at rupture or percent neck- yield stress Sy), it will return to its original shape once the load is
ing down of the original cross section. The concrete design code ACI removed. A material with high tensile stress Su will be able to absorb
318-02 Appendix D h as a good defmition of ductility: "Ductile steel ele- a large load (provided it has sufficient toughness) before it ruptures.
ment-An element with a tensile test elongation of at least 14 percent A material with high elongation at rupture, high ductility eu, will be
and reduction in area of at least 30 percent. A steel element meeting able to stretch and neck down before it ruptures. This is the case for
the requirements of ASTM A 307 shall be considered ductile."23
steel at room temperature. It is obviously not the case for glass or,
more important for our purpose, of carbon steel or low-alloy s~eel at
TABLE 2.6 Varlatlon of Yleld Stress (ksi) with Temperatura very low temperature. The ability to stretch before rupture (a high eu)
ASTM 100°F 200°F 300°F 400°F 500°F 1000°F is essential in limiting the size of the rupture.
A good example of stretch before rupture is shown in Fig. 2.25,
A516Gr.70 38 34.8 33.6 32.5 31.0 22.6 where a stainless steel instrument tube burst in a ductile manner by
A 106 Gr.B 35.0 32.1 31.0 29.9 28.5 20.8 overpressure. This is a ductile rupture, also referred to as a ductile
A213Tll 30.0 27.7 26.3 25.3 24.4 18.8 fracture. The material bulges out at one point around the circumfer-
A 240 201-1 38.0
ence stretches the surrounding metal, thins down, and fmally rup-
ture~. The shape of the ductile fracture by overpressure is sometimes
28.9 25.0 22.7
Figure 2.25 Ductilc fracture of steel tubing by overpressure.
referred to as fishmouth. By stretching, the highly strained zone

receives reinforcement from the neighboring material, which is trying
to hold it down, before it finally ruptures. This ability of ductile mate-
riais to stretch and secure reinforcement from adjacent metal is impor-
tant in reducing the risk of sudden and long brittle fractures. It is at
the basis of the metal's capacity to leak before it breaks. Figure 2.26
shows the same effect on a larger scale.
In Fig. 2.26 notice the two ends of the fracture, at the top and bot-
Figure 2.26 Ductile fishmouth fracture.
tom. The fracture surface turns at 45° from the original longitudinal
opening direction. This is common in large ductile fractures. As the
fracture progresses, the metal bulges out and then tears, which leads • Failure is caused by maximum shear, it occurs more readily in thin
us to define a ductile fracture as a fracture that occurs when the sections, with a stress nearly constant through the thickness, a con-
shear stress exceeds the shear strength of the material before the dition referred to as plane stress.
normal stress exceeds its cohesive strength. In particular, ata notch, • Failure occurs above the nil ductility transition temperature <NDT)
"the normal stress at the root of the notch will be increased in rela- of the material at its thickness.
tion to the shear stress and the [material] will be more prone to brit-
tle fracture."13 • The fracture has shear lips, a fibrous (dull) surface, and sharp edges
For a component with uniform wall thickness, it is practically protruding at 45°, with a cup-and-cone shape (Fig. 2.24).
impossible to predict where around the circumference this bulging and • 1t is more prevalent in thin sections, at relatively higher tempera-
rupture will take place, because it will depend on inevitable local tures, and at low strain rate (quasistatic loading).
changes in wall thickness and metallurgy. • The fracture is limited in size; the component tends to leak rather
A ductile fracture exhibits the following characteristics: than have a long running break.
• The component undergoes visible deformation and yielding before
fracture. 2.22 Brittle Fracture
• There is substantial plasticity; the stress has to reach the ultimate The opposite of a ductile fracture is a brittle fracture. A brittle fracture
strength of the metal Su. occurs with little deformation (little bulging in the case of overpressure),
/
Figure 2.27 Brittle fracture by overpressure of steel vessel.
Figure 2.28 Brittle fracture, flat fracture :;urface.
and can happen even below the yield stress, as illustrated in Figs. 2.27
to 2.29. The brittle fracture is sudden and flat in appearance (not a fish-
mouth), the weak spot receiving practically no help from the surround-
ing material. Unlike a ductile fracture, there is little necking-down of
the section, and no cup-and-cone shape in the fracture surface; on the
contrary, the fracture swface is flat, as shown in Fig. 2.28, quite differ-
cnt from Figs. 2.24 and 2.25. Brittle fractures can only be explained and
therefore prevented by understanding the second mechanical property
o f materiais: toughness.
A brittle fracture exhibits the following characteristics:
• Fracture occurs suddenly, with little deformation.

• Can occur ata stress below yield Sy.
• Can be caused by high constraint, such as a thick wall, with equal
principal stresses, and near constant strains through the thickness,
a condition referred to as plane strain. Figure 2.29 Br1ttle fracture of a pressure vesscl during shop
• Occurs if temperature is below the nil ductility transition tempera- hydrotest.
ture (Sec. 2.24).
• The fracture surface is flat, tends to be shiny unless covered with
• Brittle fracture is more prevalent in thick sections, at relatively lower
cotTosion products or process fluid deposits; the fracture swface has
temperatures, and at high strain rate ( dynanuc impact).
marks in the shape of">>>" referred to as chevron marks, that point
to the origin of the fracture. • It tends to cause a large break rather than a leak.
2.23 Toughness the test apparatus. The difference between the pendulum's initial
Toughness is24.25 and final heights multiplied by its weight is the energy absorbed by
the specimen, related to its Charpy toughness, measured in foot-
• The ability to absorb impact energy pound (ft·lb). Charpy V-notch toughness of a material depends on the
• The ability to hold a crack stable under load following:
Toughness depends on the following: • The test temperature

• The material composition
• The material. Cast irons have low toughness; stainless steels have
• The microstructure
high toughness. High-carbon steels have less toughness than low-
carbon steels, as illustrated in Fig. 2.30. The Charpy toughness versus temperatura curve has an S-shape
• The temperature. The lower the temperature the lower the tough- (Fig. 2.30), and the inflection point (the midpoint ofthe 8) is referred to
ness (Fig. 2.30). Note the inflection point of the S-shaped curve at as the NDT temperature or the fracture appearance transition temper-
the nil ductility transition temperature. Some materiais, such as atura (FATT). A fracture at the FATT has a 50 percent brittle appear-
austenitic stainless steels, can be tough down to cryogenic tempera- ance and 50 percent ductile appearance.
turas of near -350°F. Minimum requirements for toughness are specified in design codes
• The thickness. Thínner materiais are tougher when subject to the and !>Landard~;; fur examvle, fur ga!> pip1:üines ASME B31.8 specifies
same stress Ooad per unit area). that the all-heat average of the CVN must be sufficiently large so
that47
• The strain rate. Toughness decreases under dynamic load.
CVN > 0.0345 crL5 R 0 ·5
2.24 Charpy Toughness
Charpy V-notch toughness is measured by the strike of a swinging
pendulum against a V-notched specimen (Fig. 2.31)S The specimen
can be placed horizontally (Charpy test) or vertically (lzod test) in
I I
=º 250 ,---r ------
.i: I 0.1%C
li
(1J
200 - - - , - - -r--- ---
1
c: I
I I
-§,
::J
150 ,---r------,-- r ---~--
1 I
.9
e: 100
<ti
0.3%C
.c:
I
(.) 50 --.., ---r---~- --
I
-1 00 o 100 200 300 400 500

Temperatura, °F
Figure 2.30 'lbughness decrease with temperature. Figure 2.31 Charpy V-notch test.
where CVN = Charpy V-notch toughness at minimum operating tem-

perature, ft-lb
a = hoop stress in pipe wall due to pressure, ksi
R = pipe radius, in
For example, a 20-in API 51 X60 gas pipeline that operates ata hoop
stress of 72 percent Sy, and has a minimum operating temperature of
32°F must have a minimum Charpy V-notch toughness, at 32°F, of
CVNmin = 0.0345 X (0. 72 X 60)1.5 X (20/2)0·5 = 31 ft-lb
Not ali material specifications require toughness testing. Some do,
as a standard (mandatory) or supplementary (optional) require-
ment. 26-31 Toughness is imposed by means of Charpy testing or other
tests such as fracture toughness testing, nil ductility drop-weight test-
ing, drop weight tear testing, and dynamic tear testing.32-44
Figure 2.33 Close-up view of crack opening.
2.25 Fracture Toughness
The fracture toughness test is based on the slow quasistatic loading of
a cracked specimen (Figs. 2.32 and 2.33). This test establishes the material's elastic fracture toughness K 1c, as a function of temperature,
measured in units of ksi(in)0·5•
Because a fracture toughness test requires a sizable specimen, the
machining of a precise notch, and the formation of a crack by fatigue,
it is relatively difficult to conduct. Unlike the more conventional mate-
rial properties such as yield stress Sy and ultimate strength Su, it is
often difficult to find published values of the fracture toughness KIC for
a given material and temperature. In the absence of fracture tough-
ness tests, fracture toughness can be estimated from approximate
relationships with more comm.on material properties. For example,24.45
K - S
IC- Y
~S(CVN)
S - 0·25
y
where K 1c = linear elastic fracture toughness, ksi(in)05

Sy = material yield stress, ksi
CVN = material Charpy V-notch toughness, ft-lb
In the above example, theASME B31.8 code requirement for the min-
imum Charpy toughness ofa 20-in,API 5LX60 gas pipeline, was 31 ft-lb
at 32°F; this conesponds to an approximate fracture toughness of
Figure 2.32 Toughness testing and crack opening displace-

ment.
Krc = 60)
5
~~ 1 - 0.25 = 92 ksi-.lm"
2.26 Toughness Exemption Curve 120

I
I I I
To prevent brittle fracture, design codes require that components 100 --- -~------ -~- ----- -~- - -- ---- - :::_
- :..
-
_..:.-=-~---1
I
above a certain thickness and below a certain temperature be tested to
verify that they have a minimum levei of toughness. For ferritic pres- Ol
80
I 1
------ --
I
---.------
I
sure vessels, the threshold thickness-temperature below which tough- .!: LI-
iij O 60 ---- .-- ·~ ------~- --- - - J
I I
______ j
I
___ __ _
ness testing becomes a requirement, called the toughness test Q;ai

a.'-
I I I
o~
40 -----J------~------i
1 I ----- -
exemption curve, is provided in ASME VIII Division 1, Fig. UCS 66 for E~ _, __ I
severa! materials. 48 These curves were developed based on the follow- :l Q) 20 ---~----- - ~ ------
E a. 1
ing relationship. ·-
c E Q) o .
I
-~---- -- + ------
~- 1 I
-- - I
I
-20 r------------~ --~--- ---T------
I I
I I
-40 -- r - - - - - - - - -:- - - - - ., - -
~ , - -----T------
1
-60+--;---r-----,------.-----~------T------J
1
where K1c = fracture toughness, ksi...Jin o 0.4 1.0 2.0 3.0 4.0 5.0
Sy = yield stress, ksi Wall thickness, in
T = operating temperatw·e, °F Figure 2.34 General fot·m of toughness exemption curve for SA 515 Grade 60
To - nil ductility transition, °F (minimum operating tempcraturc vs. wall thicknessJ.
Together with the fracture toughness approximation

• It is a measure ofthe susceptibility ofthe material to cracking mech-
anisms; the harder the metal the more prone to cracking.
• It is an indirect indication of residual stresses in a weld heat-affected
where t = plate thickness, in. zone.
• It is a way of estimating ultimate strength, as was illustrated in
This leads to the toughness exemption curves T(t ) given by Table 2.5.
The most common hardness testing methods and scales are Vickers
~ = 1.7 + 1.37 tanh ( T ;6To ) hardness, Brinell hardness, Rockwell hardness, Knoop microhard-
ness, and Scleroscope hardness. There are hardness conversion tables
and equations to convert a hardness reading from one scale into
The toughness exemption curve depends on the material, but has another. 49-55
the general form of Fig. 2.34. For combinations of minimum operating
temperature and wall thickness below the toughness curve the mater- • The Rockwell hardness test is an indentation test using a calibrated
ial runs the risk of brittle fracture. In these cases, a different materi- machine to force a conical indenter into the surface. The indenter
al should be selected or the material should be procured with may be a diamond cone ot· a steel or tungsten carbide ball. The
minimum toughness at the lowest operating temperature. The mini- indentation is made in two steps with a preliminary (minor) load
muro toughness should be selected based on a fracture analysis of the and a total (major) load. The hardness number is the increase in
component, in accordance with Chap. 9. depth of the indentation from minor to major load. The nomencla-
ture for a Rockwell hardness is, for example, 60 HRC where 60 is the
hardness, HR stands for hardness Rockwell, and C refers to a dia-
2.27 Hardness
mond cone test, whereas 70 HRBW is a hardness Rockwell test (HR)
Hardness is the resistance of a metal surface to indentation. In fitness- with a ball {B) made of tungsten (W). There is a wide-ranging scale
for-service practice, hardness is of interest for three reasons: of R.ockwell hardness, from A to Y.
• The Vickers test uses a square-based pyramidal diamond to indent 22. Antaki, G.A., Piping and Pipeline Engineering, Dekker, New York.
23. ACI 318-02, Building Code Requirements for Structural Concrete, Amcrican
the surface under a predetermined force; the hardness is related to Concrete lnstitute, 2002, Detroit, MI.
the length of the diagonal of the indentation. 24. Barsom, J.M., Rolfe, S.T., Fracture and Fatigue Control in Structures, ASTM
International. West Conshohocken, PA.
• The Brinell test is conducted with a tungsten carbide ball to obtain 25. Boyer, H.E., Ed., ASM, Atlas o{ Stress-Stram Curves, ASM International, Materiais
a Brinell hardness labeled HBW (older tests used a stee] ball, in Park, OH.
which case the hardness is labeled HB or HBS). 26. ASTM A 573, Standard Specification for Structural Carbon Steel Plates oflmproued
7bughness
• The Knoop hardness is a microindentation, on the order of microns, 27. ASTM A 334 Standard Specification for Seamless and Welded Carbon andAlloy-Steel
Thbes for Low-Temperature Seruice, ASTM lnternational, West Conshohocken, PA.
that measures the microhardness of a very thin or very small part 28. ASTM A 333, Standard Specification for Seamless and Welded Steel Pipe for Low-
and coating. Temperature Seruice, ASTM International, West Conshohocken, PA.
29. ASTM A 420, Standard Specification for Pipmg Fittings o{ Wrought Carbon Steel
• The Scleroscope test measures hardness through the height of and Alloy Steel for Low-Temperature Service, ASTM International West
rebound of a diamond-tipped hammer. Conshohocken, PA.
30. ASTM A 350, Standard Specificaticn for Carbon and Low-Alloy Steel Forgings,
Requírmg Notch 7bughness Testing for Piping Components, ASTM International,
References West Conshohocken, PA.
31. ASTM A 352, Standard Specification for Steel Castings, Ferritic and Martcnsitic, for
1. The Metais Black Book, Volume 1, Ferrous Metals, CASTI Publishing, Alberta, Pressure-Containing Parts, Suitable for Low-Temperature Seruice, ASTM
Canada. International, West Conshohocken, PA.
2. The Metals Red Book, Volume 2, Nonferrous Metais, CASTI Publishing, Alberta, 32. ASTM E 208, Standard Test Method for Conduclmg Drop-Weight Test to Determine
Canada. Nil·Ductility Transition Temperature of Ferritic Steels, ASTM International, Wcst
3. ASM Handbook, ASM lntemational, Materiais Park, OH. Conshohockcn, PA.
4. The Nickel Institute, 'lbronto, Ontario, Canada. 33. ASTM E 436, Standard Test Method for Drop-Weight Tear Thsts o{ Femtcc Steels,
5. ASTM A 941, Terminology Relating to Steel, Stainless Steel, Related Alloys, and ASTM lntemational, West Conshohocken. PA.
Ferroalloys, ASTM Intemational, West Conshohocken, PA. 34. ASTM E 604, D.vnamic Tear Testing o{Metallic Materiais, ASTM lntemational, West
6. ASTMA247, Standard Method for Evaluating the Microstructure o{Graphite in lron Conshohocken, PA.
Castings, ASTM lnternational, West Conshohocken, PA. 35. ASTM E 1823, Standard Terminology Relatil1g to Fatigue and Fracture Testing,
7. ASTM A 126, Starulard Specification for Gray lron Castings for Valves, Flanges, and ASTM Intcrnational, West Conshohocken, PA.
Pipe Fittings, ASTM International, West Conshohocken, PA. 36. ASTM E 399, Standard Test Method for P/anc-Strain Fracture 7bughness o{ Metallic
8. ASTM A 370, Standard Test Methods and Defi.nitions for Mechanical Materiais, ASTM International, West Conshohocken, PA.
Testing of Steel Products, ASTM International, ASTM International, West 37. ASTM E 436, Standard Test Method for Drop Weight Thar 'Thsts o{ Ferritic Steels,
Conshohocken, PA. ASTM Intcrnational, West Conshohocken, PA.
9. Metallic Materiais and Elements for Aerospace Vehicle Structures, U.S. Department 38. ASTM E 604, Dynamic Tear Testirtg ofMetallic Materiais, ASTM lntcrnational, WeHt
of Defense Handbook, MIL-HDBK-5H, 1 December, 1998. Conshohocken, PA.
10. AB/Coi!Tech, Sweden. 39. ASTM E 812, Standard Test Method for Crcu·k Strength o{ Slow-Bend Precracked
11. Van Droffelaar, H., Atkinson, J.T.N., Corrosion and lts Control. An lntroductian to Charp_y Specimens o{ High-Strength Metallic Materiais, ASTM lnternatíonal, West
the Subject, NACE lnternational, Houston, TX. Conshohocken, PA.
12. Chasis, D.A., Plastic Piping S_vstems, Industrial Press Inc. 40. ASTM E 1221, Standard Test Method for Determining Plane·Strain Crack·Arrest
13. SPI Society ofthe Plastics Industry, Inc., Fiberglass Pipe Handbook, Fiberglass Pipe Fracture Toughness, Kla, o{ Ferrític Steels, ASTM lnternational, Wcst
lnstitute, 2d ed., 1992, New York. Conshohockcn, PA.
14. UniBell, Handbook o{ PVC Pipe Design and Construction, The UniBell PVC Pipe 41. ASTM E 1290, Standard Test Method for Crack·TI.p Opening Displacemenf (CTODJ
Association, Dallas, TX. Fracture 1bughness Measurement, ASTM lnternational, West Conshohocken, PA.
15. Kiefner & Associates, Worthington, OH. 42. ASTM E 1304, Standard Test Method for Plane·Sfrain (Chevron-Notclzi Fracture
16. WRC Bulletin 452, Recommended Practices for Local Heatmg o{ Welds in Pressure 1bughne11s of Metallic Material, ASTM Internotional, West Conshohocken. PA.
Vessels, Pressure Vessel Research Council, June 2000. 43. ASTM E 1737, Stamlard Test Method for J Integral Characteri.zation o{ Fracture
17. ASTM E 6, Startdard Terminology Relating to Methods o{Mechanical Testing, ASTM Touglmess <Discontinued 1998; replaced by E1820J, ASTM Intcrnational, West
Intemational, West Conshohocken, PA. Conshohocken, PA.
18. ASTM E 8, Standard Test Methods for Tension Testing of Metallic Materiais, ASTM 44. ASTM E 11:!20, Standard Test Method for Measurement o{Fracture 7bughness, ASTM
lnternational, West Conshohocken, PA. International, WcRt Conshohocken, PA.
19. Landon, J., Chu:ago Bridge and lron, CBI, Chicago. 45. Doggett, M., California State University, Fresno, Department of Industnal
20. Haggag, F.M., In-Service Nondestructiue Measurements o{ Stress·Strain Curves and Technology.
Fracture Toughness o{ Oil and Gas Pipelines: E:mmples o{ Fitness-for-Purpose 46. API RP 579, Fitness-for-Service, American Petroleum Institute, Washington, DC.
Applications, Advanced Technology Corporation, Oak Ridge, TN. 47. ASME B31.8 Gas Transmission and Di.~tribution Piping, American Socicty of
21. Haggag, F.M.,In·Situ Nondestructive Measurements o{ Ke.v Mechamcal Properties o{ Mechanical Engineers. New York.
Pressure Véssels Using lnnovatiue Stress·Strain Microprobe (SSM) Technology, 48. ASME Boiler and Pressure Vessel Code, Section VIII, Rules for Construction o{
DOE/ER/82115-1, March 5, 1997. Pressure Vesse/s, American Society of Mechanical Engineers, New York.
66 Chapter Two
Chapter
3
49. ASTM E 18, Standard Thst Methods for Rockwell Hardness and Rockwell Superficial
Hardness of Metallic Materiais, ASTM lntemational, West Conshohocken, PA.
50. ASTM E 140. Standard Hardness Conversion Tables for Metais Relationship Among
Brinell Hardness, Vickers Hardness, Rockwell Hardness, Superficial Hardness,
Knoop Hardness, and Scleroscope Hardness, ASTM International, West
Conshohocken, PA.
51. ASTM E 10, Test Method for Brinell Hardness of Metalltc Materiais, ASTM
lntemational, West Conshohocken, PA.
52. ASTM E 18, Thst Method for Rocku.-ell Hardness and Rockwell Superficial Hardness
of Metallic Materiais, ASTM Jntemational, West Conshohocken, PA.
53. ASTM E 92, Tl>st Method for Vickers Hardness of Metalltc Materiais, ASTM
Design
54 ASTM E 384, Test Method for Microhardness of Materiais, ASTM Intemational,
West. Con~hohocken, PA.
55. ASTM E 448, Practice for Scleroscope Hardness Testing of Metallic, ASTM
International. West Conshohocken, PA.
56. ASTM E 112, Standard Thst Methods for Determining Average Grain Size, ASTM
Intemational, West Conshohocken, PA.
57. ASTM B 248, Standard Specification for General Requirements for Wrought Copper
and Copper-Alloy Plate, Sheet, Strip, and Rolled Bar, ASTM lnternational, West
Conshohocken,PA.
58. ASTM B 152, Standard Speci(lcation for Copper Sheet, Stríp, Plate, and Rolled Bar,
ASTM lnternat10nal, West <.:onshohocken, PA.
59 ASTM B 283, Standard Specification for Copper and Copper-Alloy Die Forgings, 3.1 Basic Design and Detalled Deslgn
ASTM International, West Conshohocken, PA.
60. ASTM B 75, Standard Specification for Seamless Copper Thbe, ASTM lnternational, Basic design is the design of the process and the system logic, includ-
West Conshohocken, PA. ing the operating parameters, process controls, output, flow, thermohy-
61. ASTM B 88, Standard Specification for Seamless Copper Water Thbe, ASTM
International, West Conshohocken, PA. draulics, and safety requirements. The basic design sets the plant, unit
62. ASTM B 359, Standard Specificat10n for Copper and Copper-Alloy Seamless and system layout, and equipment size. For example, the size of a heat
Condenser and Heat Exchanger Thbes with Integral Fins, ASTM lnternational, West exchanger (diameter, length, nurnber of tubes) is set by the flow rates
Conshohocken. PA.
63. ASTM B 43. Standard Specificatíon for Seamless Red Brass Pipe, Standard Sizes. and required heat transfer capacity; the size (diameter) of a pipeline is
ASTM lnternational, West Conshohocken, PA. set by flow rate, through-put and pressure drop; and the size of a stor-
64. ASTM B 188, Standard Specífication for Seamless Copper Bus Pipe and Thbe, ASTM age tank (height, diarneter) is set by its expected storage capacity. Basic
65. ASTM B 608, Standard Specification for Welded Copper-Alloy Pipe, AST~i design also includes accident analysis and overpressure protection
International, West Conshohocken, PA. (safety and relief devices selection and sizing).
66. ASTM A 336, Standard Specificalion for Seamless Ferritic Allay-Steel Pípe for High- Once the basic design has been set, the detailed design phase follows.
Tl>mperature Seru1ce, ASTM International, West Conshohocken, PA.
67 Viswanathan, R., Damage Mechanisms and Life Assessment of High-Thmperature The detailed design is the design of equiprnent and components for
Comporumts, ASM International, Metais Park, OH. strength and integrity; it addresses rninimum required thickness, stiff-
68. Thielsch EnFlineering, Cranston, RI. eners, reinforcements, and supports. Detailed design activities follow
well-established design codes.
3.2 Design Codes

Vessels, piping, and tanks are sized and laid out following the rules of
design and construction codes and standards. The design codes include
the following.
The ASME Boiler and Pressure Vessel Code comprises 12 sections
labeled I to XII. For simplicity, when referring to the Boiler and Pressure
Vessel Code, we simply note ASME followed by the code section in
Roman numerais and, if applicable, the Division; for example, ASME
67
68 Chapter Three Deslgn 69
VIII Div.1 refers to ASME Boiler and PressW"e Vessel Code, Section VIII, • ASME B31.9, Building Services Piping. Low pressure steam and
Division 1. The 12 ASME Boiler and PressW"e Vessel Code sections are water distribution.
as follows. • ASME B31.11, Slurry Transportation Piping. Mining, slurries, sus-
• Section I, Rules for Construction of Power Boilers pended solids transport, and the like.
• Section li, Materiais The API standards for design of storage tanks are as follows.
• Section 111, Rules for Construction of Nuclear Facility Components • API 620, Design and Construction of Large, Welded, Low Pressure
• Section IV, Rules for Construction of Heating Boilers Storage Tanks
• Section V, Nondestructive Examination • API 650, Welded Steel Tanks for Oi/ Storage
• Section VI, Recommended Rules for the Care and Operatwn of
The AWWA standard for design of water storage tanks is the following:
Heating Boilers
• Section VII, Recommended Guidelines for the Care of Power Boilers • AWWA D 100, Welded Steel Tanks for Water Storage
• Section VIII, Rules for Construction of Pressure Vessels The design equations used in sizing and layout include safety mar-
• Section IX, Welding and Brazing Qualification gins and corrosion allowances that will permit the component to oper-
• Section X, Fiber-Reinforced Plastic Pressure Vessels ate safely during its design life, but only if the in-service loads and
degradation mechanisms have been conectly predicted and accounted
• Section XI, Rules for In-Service lnspection of Nuclear Power Plants for at the design stage.
• Section XII, Transport Vessels
The ASME B31 Pressure Piping Code comprises several book sec- 3.3 Oesign Minimum Wall t mln
tions.1 Each design code contains formulas to calculate the minimum wall
• ASME B31.1, Power Piping. Piping systems in fossil-fueled power thickness required, given the following parameters:
plant, nuclear-powered plant with a construction permit predating • Material
1969 (B31.7 for 1969-197l,and ASME III post-1971).
• Design pressure and temperature
• ASME B31.3, Process Piping. This is the broadest code section. It cov-
ers basically ali pressW'e piping applications not explicitly in the • Applied loads
scope of the other ASME B31 sections. This scope includes chemical • Component shape
processes, hydrocarbons (refuúng and petrochemicals), the making of • Weld joints efficiency
chemical products, pulp and paper, pharmaceuticals, dye and color-
ings, food processing, laboratories, offshore platform separation of oil The minimum wall thickness required by these codes is referred to
and gas, and so on. as tmin· For example, the ASME VIII Div. 1 design equation for a pres-
• ASME B31.4, Pipeline Transportation Systems for Liquid sure vessel may show that a vessel needs to be 0.40-in thick to sus-
Hydrocarbons and Other Liquids. Upstream liquid gathering !ines and tain the design pressure. The minimum required thickness tmin
tank farms, and downstream transport and distribution of hazardous includes an inherent ASME VIII Div.1 design margin of 3.5 against
liquids (refined hydrocarbon products, liquid fuels, carbon dioxide). rupture.
• ASME B31.5, Refrigeration Piping. Heating ventilation and air con-
ditioning in industrial applications. 3.4 Future Corrosion Allowance FCA
• ASME B31.8, Gas Transmission and Distribution Piping. upstream To the minimum design thickness required by code, the designer
gathering lines, onshore and offshore, downstream transport should add a future corrosion allowance, consistent with two parame-
pipelines, and distribution piping. ters: the expected corrosion rate and the design life
70 ChapterThree Deslgn 71
FCA= CR X T,
where FCA - future corrosion allowance, in
CR = corrosion rate, in/yr
T = design life, yr
The corrosion rate and the design life are not found in design codes;
instead they are the responsibility of the owner or the designer. They
can be obtained from several sources.
• Ideally, prior operating and inspection experience with similar sys-

tems
• Company procedures and specifications
• Laboratory simulation
• Corrosion coupons in service
• Industry standards and guides such as those published by NACE,
ASM, API, and EPRI.
• Vendor catalogs (particularly for nonmetallic materials and trims) Figure 3.1 Example of"live load" on pipeline.
• Corrosion textbooks
For example, the designer ofthe above 0.4-in-thick vessel may choose 3.5 Loads, Stresses, and Strains
to add 0.10 in as a future corrosion allowance for 20 yr of service at a In service, components are subject to loads (forces and moments).
projected corrosion rate of 5 mils/yr (0.005 X 20 yr = 0.10 in); the ves- Loads arise from normal operation or from abnormal conditions.
sel is therefore fabricated with a wall thickness of 0.40 + 0.10 = 0.50
in. If the corrosion rate tums out to be indeed 5 mils/yr, then the ves- • Normal operating loads include the component's weight, internai or
sel will end its design life, after 20 yr, with a remaining wall of 0.40 externai pressure, temperature, normal variations in flow rate, safe-
in, and the intended design margin to failure of 3.5. But if the actual ty and relief valve discharge, low amplitude vibration, normal wind
corrosion rate tums out to be double the design projection, or 10 outdoors, normal waves and currents on subsea pipelines, and resid-
mils/yr, then, after 20 years, the remaining wall will be 0.50- (20 X ual stresses from fabrication and welding. Figure 3.1 illustrates
0.010) = 0.3 in and the margin to failure at end of life will be 3.5 X that, in some cases, normal loads may also include "live loads."
(0.3/0.4) = 2.6. • Abnormalloads include pressure transients from liquid hammer (Fig.
As we show in later chapters, there are other considerations that 3.2); two-phase flow transients; temperatures and pressures outside
come into play in fitness-for-service and margin assessments, but this the design range; rapid temperature fluctuations; high amplitude
simple example illustrates the importance of design margins and ini- vibration; extreme winds, waves, or currents; large ground settle-
tial corrosion allowance in fitness-for-service. ments; and earthquakes (Fig. 3.3).
A fitness-for-service assessment practically always starts with the
calculation o f the minimum wall thickness required by the design code Most loads are not constant, but fluctuate during service, with the
tm;n, and then proceeds to evaluate to what extent the degraded condi- possible exception of the component's weight, and even the weight
tion encroaches on this design thickness tmin. So, understanding design changes depending on whether the component is empty, full of water
rules and design margins, the subject of this chapter, is essential in for hydrotest, or full or partially full of liquid in service. Some fluctu-
fi tness-for-service assessments. ations are gradual, such as the change of pressure and temperature at
startups and shutdowns. Other loads fluctuate continuously, such as

vibration in service. Finally, other loads have few but severe cycles,
such as earthquake or transient hammer loads.
3.6 Applied Loads and Residual Stresses

Some loads are imposed directly as forces and moments, the simplest
case being weight or internai pressure. Other loads result from
imposed movements Cdisplacements or rotations), for example, move-
ments caused by thermal expansion of a tall distillation column, or soil
settlement beneath a storage tank.
Residual stresses are not caused by externally applied loads or
movements, instead they are built in during mill and shop fabrication,
and field erection. Residual stresses are the most difficult to compre-
hend because they are not as intuitively evident as stresses due to
imposed loads or movements. A detailed discussion of residual stress-
es is provided in Chapter 4, but at this point we can mention some
examples of residual stresses.
Figure 3.2 Pipeline burst caused • Stresses from cold bending in the shop or in the field.
by liquid hammer in corroded
section. • Cold springing of short sections by using "come-alongs" (crank and
cbains) to hold the two ends in place while welding or bolting. Ifthe
Figure 3.4 Line springs open when flange bolts are

Figure 3.3 Storage tank buckles during earthquake. removed.
74 ChapterThree Design 75
'tzy
Figure 3.5 Stress distribution a t

a point .
Figure 3.6 Bending stress distribution.
joint is reopened, the two sides of the pipe will spring back open w
(Fig. 3.4). l l l l l l l l 1 l l
• Weld contraction stresses. b a
am L
/747
3.7 General Stresses
Loads (forces and moments) cause stresses and strains at each point of Figure 3.7 Weight load in pipe span.
the component. By the elassic theory of elasticity (strength of materiais)
we know that if instead of a point we consider a very small cub€ of
metal, each of the six faces of the cube will see three stresses, one nor-
mal stress <J perpendicular to the face, and two shear stresses 't in the where M = maximum bending moment in pipe section, in-lb
plane of the face (Fig. 3.5). w = linear weight of pipe, lb/in
The magnitude of each stress component depends on the applied L = length of pipe span between supports, in
load (forces and moments), the shape ofthe component, its mechanical
The corresponding maximum bending stress is
properties, and its restraints.
(j
M
=-
wL2 1
(j ----
3.8 Example: Bending Stress L Z L - 10 Z
For simple loads applied to components of simple shape, the resulting where a L = maximum longitudinal stress dueto bending, psi, and Z =
stress distribution in the component can be predicted by a formula, a pipe cross-section modulus, in3.
"closed-form solution." For example, the stress through the cross sec-
tion of a pipe span of section modulus Z, subject to an end moment M,
as shown in Fig. 3.6, varies linearly from <Js. = + MIZ (tension) at the
outer fiber in tension, through zero at the side walls in the middle of where I = moment of inertia of pipe cross section, in4, and R = p1pe
the pipe, to <Js = -MIZ compression at the opposite outer fiber. radius, in.
The maximum bending moment due to weight in a pipe span (Fig.
3.7) is For thin wall cylinders
wL2 Z = O0982 D
4
- d
4
M =.- . D
10
where D = outside diameter, in and d = inside diameter, in.

For example, for a 20-ft (240-in) long horizontal span of 14-in
schedule 30 pipe, with a weight of w = 55 lb/ft, and a section mod-
ulus Z = 53 in3, in gas service, the maximum bending stress is
2
a L = (55 I 12) x (240) 1 "" 500 psi
10 53
At this point we only note that we were able to calculate the stress
by a simple formula, and that it is small compared, for example, to the
yield stress of a Grade B carbon steel which is around 35,000 psi.
3.9 Pressure Stress

The stresses in a thin wall cylindrical shell subject to internai pressure
P, away from stress risers (also referred to as "structural discontinu-
ities") are illustrated in Fig. 3.8 and are equal to Figure 3.8 Stresses due to inter-
nai pressure.
PD
(jL=-
4t
(jr.OD = P. ar.ID = P = 500 psi
where ah = hoop (circumferential) stress, psi a,.,oo = Pe = Opsig =O psi

crc = circumferential (hoop) stress, psi
<JL = longitudinal stress, psi
3.11 Wall Thickness Selection
cr,.,m = radial stress at inner diameter, psi
cr,.,on = radial stress at outer diameter, psi In every equipment and pipe, at beginning of design life, we need to
P = internai pressure, psi have suffi.cient wall thickness to sustain the design (maximum) pres-
Pe = externai pressure, psi sure tmin, and to aHow for future corrosion. The required thickness is
t = wall thickness, in therefore
D = diameter, in trequire<l = tmin + FCA,
where required wall thickness at beginning of design life, in
trequired =
3.1 O Pressure Stress Example
= minimum wall thickness required by design code, in
tmin
The stresses in a 14-in schedule 30 pipe, subject to an internai pres- FCA = future corrosion allowance, in
sure of 500 psi, and no other load, are
Before ordering, a fabrication tolerance must also be added, so that
PD 500x14 .
a 1 =ac = - = = 9333 ps1 tordered = (1 + tol) trequired
' 2t 2x0.375
aL = PD = ac =4667 psi where tol = fabrication tolerance on wall thickness, from the material
4t 2 specification.
For example, if given the operating and design loads for a 6-in car- With a corrosion allowance of 3 mils/yr, the minimum required thick-
bon steel pipe, the code-required wall thickness for the pipe is tmin ness for 40 yr of service is
0.10 in, and the design corrosion rate is 3 mpy (3 mils per year), then
the thickness required for 20 years of service is trequired = 1.5 in + 0.003 in X 40 = 1.62 in
frequired = 0.10 in + 0.003 in X 20 = 0.16 in ASTM A 335 pennits defects up to 12.5 percent of the wall, but not
encroaching on the minimum required wall thickness. Adding a fabri-
If the pipe material selected is seamless carbon steel ASTM A 106, cation thickness undertolerance of 12.5 percent, the ordered wall
the ASTM material specification permits the pipe mil1 to deliver pipe thickness is
with a 12.5 percent underthickness. The tbickness to be ordered
should therefore be, as a minimum tordcred = 1.62 in X 1.125 = 1.8 in
fordcrecl - (1 + 0.125) 0.16 = 0.18 in The final step is to select the closest, and larger, commercial size. In
this case, the order could be for a commercial schedule 120 with a wall
The final step is to select the closest, and larger, commercial size. thickness of 1.812 in. Note that i f the actual pipe has its full nominal
Commercial sizes of metallic piping are specified in the following stan- schedule 120 wall t hickness, the operating hoop stress is
dards:
_PD_ = 520 x 24 = 3444 si
• ASME B36.10, Welded and Seamless Wrought Steel Pipe 2t 2x 1.812 P
• ASME B 36.19, Stainless Steel Pipe
Because this is a carbon steel pipe, ASME B36.10 applies. The clos-
est, and larger, commercial size for a 6-in pipe with a 0.18-in wall is 3.13 Butt-Welded Fittings
schedule 40 which has a standard wall of 0.280 in. In accordance with ASME B16.9,4 butt-welded fittings are ordered to
the same schedule as the ptpe, in the case of the example in Section
3.12 the butt welded fitting (elbow, tee, reducer, cap) would be ordered
3.12 Fossil Power Plant Example as schedule 120.
A 24-m steam line in a fossil power plant has a design pressure of 520
psi at 1100 ,F. The material is ASTM A 3352 Grade P22, a 2XCr-1Mo 3.14 Flanges
steel alloy, with the following specified chemical composition: C =
0.05-0.15, Mn = 0.30-0.60, P = 0.025 max, S = 0.025 max, Si 0.50 According to ASME B16.5,5 a forging class compatible with an ASTM
max, Cr = 1.90--2.60, and Mo= 0.87- 1.13. The pipe was tempered at A 335 Grade P22, a 2XCr-1Mo steel pipe is an ASTM A 182 Grade 22
1250 F. The ASME B31.1 allowable stress for this material is 3.8 ksi for ging.s The material belongs to ASME Bl 6.5 Group 1.10. The
at 1100°F. ASME B16.5 pressure class that accommodates a design pressure of
Minimum wall required by the ASME B3Ll code3 520 psi at 1100°F for ASME B 16.5, Group 1.10, is Class 1500 (565 psi
at 1100°F ).
PD t . = 520 x 24 l.S in
to= --
mtn 2(SE + Py) mm 2(3800 X 1 + 520 X 0.7) 3.15 Socket and Threaded Fittings
where P- design pressure, psi The pressure class of socket-welded and threaded fittings, such as an
D = outside pipe diameter, in instrument tap on lhe 24-in line, is based on the wall thickness of the
S - code allowable stress for the material at operating tem- matching small bore pipe, in accordance with ASME B16.11.7 For
perature, psi example, for a l-in branch line with a socket-welded balf-coupling, the
E '- weld joint efficiency factor actual branch pipe diameter is 1.315 in. The wall thickness of the
y = temperature correction factor, defined in the B31 code branch pipe is
r
TABLE 3.1 Pressure Ratlng Based on Plpe Schedule
Equivalent pipe schedule Equivalent pipe schedule

Pressure class for threaded joint for socket weld
2000 80
3000 160 80
6000 xxs 160
9000 xxs
t . = PD 520 x 1.315 = O.OS2 in

mm 2<SE + Py) 2(3800 X 1 + 520 X O. 7)
After corrosion allowance and thickness tolerance, the selected l -in

pipe schedule is schedule 40, with a nominal thickness of0.133 in. Thtl
pressure ra ting of the half-coupling can now be selected, based on the
corresponding pipe schedule, following ASME B16.11, and presented,
in part, in Table 3.1. In this example, the schedule 40 branch pipe
would correspond to a Class 2000 threaded joint ora class 3000 sock- Figure 3.9 Vessel example.
et-weld fitting.
3.16 Specialty Fittings and Components

Specialty products (joints, welded or mechanical fittings, hoses, spe- The hoop stress in the cylindrical shell is
cialty components) are typically procured to specific model numbers
according to a manufacturer's catalogue and specifications. They are _ PD _ 300x(4x12) _ .
not standard ASME B16 fittings. Manufacturers design, proof test, (1"h - 2t - 2x0.375 - 19' 200 psl
and rate their products following the rules for "unlisted components"
of the design and construction codes. Proof testing is typically con- The longitudinal stress in t he cylindrical shell is
ducted on production prototypes and must achieve a safety margin
against leakage or rupture in the range of 3 to 5, depending on the 19200
= 9600 psi
applicable code, with temperature correction factors. 2
The shell will burst when the hoop stress in the shell reaches the
3.17 Vess·el Example ultimate strength o f the material, or
Apressure vessel has a diameter of 4ft and a wall thickness of 0.375
PD -S
in (Fig. 3.9). The vessel operates at 300 psi and 180°F. What is the 2t - u
hoop stress in the cylindrical shell? What is the longitudinal stress? If
the vessel is made of carbon steel with an ultimate strength of 60 ksi, Px(4x12) = 60,000
at approximately what pressure will it burst? What will be the shape 2x0.375
o f the fracture? P =937psi
In addition to code design rules (stress or strain limits), limits on defor-

mations are applied by designers as a matter of good practice. For exam-
ple, the sag in the middle of a pipeline span should not cause liquid to be
trapped at midspan and should not overcome the line slope.
The margin between the applied stresses and the yield stress or the
ultimate strength varies from code to code, as becomes evident when
we review the design formulas, later in this chapter.
Design of static equipment for normal operating loads is based on
elastic stress analysis, in which stresses are computed and limited to
a fraction of yield. Design for extreme loads, such as earthquakes, per-
mits plastic deformations in certain cases. In these cases the design is
based on plastic analysis or on simplified equivalent elastic stress
analysis.
Fatigue design can be implicit or explicit.
• Explicit fatigue analysis is an integral part of piping and pipeline
design in accordance with ASME B31. It is also addressed explicitly
in the pressure vessel design rules of ASME VIII Div.2.
• Fatigue is not addressed explicitly in the pressure vessel design
rules of ASME VIII Div.l.
Fig ure 3.10 Overpressure burst
of ductile cylinder.
If a vessel is subject to significant fatigue loads in service it should
be designed to the rules of ASME VIII Division 2, or the rules of ASME
VIII Division 1 should be supplemented to address fatigue.
This calculated burst pressure is an approximate prediction because

3.19 Design Pressure
it does not account for the plastic bulging that takes place right before
burst. This calculation also presumes that there is no fabrication flaw, The minimum design wall thickness is calculated on the basis of the
no corrosion, and that the other components of the vessels (heads, noz- system design pressure, not its normal operating pressure. When the
zles, etc.) will not fail first. system is protected against overpressure by safety or relief devices
Because 0.375-in thick carbon steel at 180°F is ductile, the rupture (valves or rupture discs), then the design pressure is the pressure set
will be preceded by an outward bulge of the vessel shell. Because the point on the relief device, plus any hydrostatic head dueto differences
hoop (circumferential) stress is twice the longitudinal stress, the Larg- in elevation (Fig. 3.11).
er stress (hoop) will tend to split the shelllongitudinally (Fig. 3.10). Ifthe system is not protected by a relief device, then the design pres-
sure is the maximum credible overpressure that may occur in the sys-
3.18 Design Principies
tem. It may be the deadhead of a centrifugal pump running against a
downstream valve mistakenly closed. It can be difficult to determine
The rules of design codes are intended to maintain the stresses in a the design pressure of a system that is not protected by a relief device,
component to a fraction of the material's yield stress and ultimate because it becomes necessary to consider all the credible "what if' sce-
strength. narios that could lead to overpressure.
Design rules for vessels, tanks, piping systems, and pipelines do not The explosion of an oil-gas separator illustrates this point (Figs. 3.12
directly limit the applied loads (forces and moments) or deformations and 3.13). The separator was designed and normally operated at
(translations and rotations). Instead, they limit stresses or, in the case atmospheric pressure, and for that reason it had no pressure-relieving
of plastic design and high-temperature creep design, they limit strains. device. Its design pressure was atmospheric pressure. During an
Figure 3.12 Overpressure exploslon and fire of vertical tank.•
Figure 3.11 Safety relief valves set system design pressure.
unusual system startup, an error in valve alignment diverted pressur-

ized gas to the vessel, causing the overpressure and subsequent explo-
sion. The design pressure of the separator did not account for the
uwhat i f' of valve misalignment during startup.
3.20 Vessel Cylindrical Shell

The minimum wall thickness of the cylindrical shell of an ASME VIII
Div.1 vessel, according to ASME VIII article UG-279 is
PR
tmin = -:-(s-=-E=-----o-.6-=P-) Figure 3.13 Overpressure explosion.A
where tmin minimum wall of the cylindrical shell, in

p design pressure, psi The vessel material is SA 515-70 carbon steel plate, with an ASME
R inside radius, in VIII Div.1 allowable stress S 17500 psi; this value is obtained from
s allowable stress, psi
ASME li Part D.
E weld joint efficiency factor
During fabrication, the vessel welds are double-welded (butt-welded
Example. The design pressure of a vessel is P = 100 psi and its from inside and out) and they are examined by spot radiography,
operating temperature is ambient 70°F. E= 0.85.
TABLE 3.2 Example of Weld Joint Efflclency

(ASME VIII Dlv.1, UW-12)
Weld type 100%RT SpotRT NoRT
Single-welded butt 0.90 0.80 0.65

Double-welded butt 1.0 0.85 0.70
RT = radiographic testing
Spot = defined in ASME VIII, for example, 1 radiography every
50 ft of weldment
The pressure vessel has a radius R = 48 in (diameter D = 96 in).

What is the ASME VIII Div.1 required wall thickness of the cylin-
drical shell?
Figure 3.14 Spherical vessels.
. = PR = 150 X 48 = 0.49 in
tmm SE- 0.6P 175()0 X 0.85-0.6 X 100
Therefore, the minirnum wall thickness required by the ASME VIII

Division 1 code for this vessel is 0.48 in, without a corrosion allowance.
This minimum code required thickness is referred to as tmin in fitness- 3.22 Elliptical Head
for-service assessments.
The minimum required wall thickness of an elliptical head, shown in
Figs. 3.15 and the top of 3.16, is9
3.21 Spherical or Hemispherical Head
PDK
The minimum required wall thickness of a spherical head, shown in t .
mm
= --=----=-
2SE-0.2P
Fig. 3.14, is9
where tmin = minirnum wall of head, in
PR P design pressure, psi
=
tmin = 2SE- 0.2P Rc inside radius, in
=
S = allowable stress, psi
where tmin = minimum wall of head, in E = weld joint efficiency factor
P = design pressure, psi K = (2 + 2Rcu)/6
R = inside radius, in
Reu = ratio of major-to-minor axis of elliptical head = B/A = 2
for 2:1 head (B = 2, A = 1)
E = weld joint efficiency factor
Example: The thickness of an elliptical head on the same vessel,
Example. The thickness of a spherical head on the same vessel as in
with E = 0.85, and Reu = 2:1 shape (K = 1) is
Sec. 3.20 is
t. -
_ 150 X 48 _Ü .
- .24 m
. = 150x96 =0.4·g in
mm 2 x 17500 x 0.85-0.2x 150 tmm 2 X 17500 X 0.85-0.2 X 150
88 Chapter Three Design 89
3.23 Torispherical Head

This awkward name, also called "flanged and dished," refers to a head
that is comprised of a toruslike bottom and a spherical top, as shown
in Fig. 3.16 at bottom.
The minimum required wall thickness of a torispherical head, as
shown in Fig. 3.17, is9
PRM
tmin = -2-,-S....,.E---0-.2-P.,...
where tmin = minimum wall of head, in

P = design pressure, psi
R = inside crown radius of spherical section, in
M = (3 + ..JR/r)/4
r = imüue k.uuck.le radius of iori:spherical head, in
3.24 Flat Head

Figure 3.15 Elliptical head.
The minimum required wall thickness of a flat head, as shown in Fig.
3.18, is9
t=d~CP
SE
----- - - --- ------r~-
-------- - --------~ - ---- ---·-·-·-·--
I
0 = 28
·-·-·-·----~·-·-·
o
Figure 3.17 Torispherical head.
Figure 3.16 Elliptical head at w p, torispherical head a t
botwm.
90 Chapter Three Desig n 91
3.26 Plant Piping- ASME 831 .3

First, the wall thickness is calculated for the straight pipe (Fig. 3.19),10
t- PD
2CSE+Py)
where P = design pressure, psi

D = outside diameter, in
E = longitudinal weld joint efficiency factor
t = wall thickness, in
S = stress allowable, psi
Figure 3.18 Flat head.
y = temperature dependent factor; for example, y = 0.4, for
austenitic steel below 1100°F
The joint efficiency factor E is a measure of the reliability of the mill

where t = minimum wall, in seam weld of the pipe, and is provided in ASME B31.3. Some typical
C = O.33 t,lts values include:
tr = minimum required thickness of seamless shell, in • Seamless E= 1.0
ts = actual thickness of shell, exclusive of corrosion
a llowance, in • Furnace butt weld E= 0.6
S maximum allowable stress, psi • Electric fusion are weld E= 0.8
E weld joint efficiency factor
d = shell diameter
Example. The thickness of a flat head on the same vessel, with

E = 1.0, tr = 0.41 in, fs = 0.60 in, and d = 96 in is
t,mn = 96x
0.33x(0.41/0.60)x150 =4 _2 in
17,500xl.O
3.25 Comparison
In summary for this vessel, the minimum required t hickness of differ-
ent parts would have to be
• Sphelical head = 0.24 in thick

• Cylindrical shell = 0.49 in thick
• Elliptical head = 0.49 in thick
• Flat head = 4.20 in thick
These results are consistent with the strength principie: In resisting

pressure, the spherical shape is strongest, the flat shape is weakest. Figure 3.19 Plant piping is designed for pressure and moment loads.
design life of the system. The corrosion allowance would have intro-
duced, for the first time in the design process, the need to establish a
design life.
The next step is to select butt-welded fitting thicknesses. This is
done simply, in accordance with ASME B16.9, by specifying the fittings
to the same schedule as the pipe. The elbow will therefore be ASME
B16.9 schedule 20 stainless steel ASTM A 182 type 304L.
Finally, the end flanges will be selected in accordance with ASME
B16.5. The flange is a forging made to ASTM A 182 Gr. F304L. This
material is assigned to Group 2.3 in ASME B16.5. The ASME B16.5
pressure rating for a Group 2.3 flange that can sustain 370 psi at
850°F is Class 400.
3.27 Plant Piping Moment Stress

Having established the wall thickness and fitting classes, the next
Figure 3.20 Jllgh-tcmperatm·c pipe. step is to veri(y the design for "sustained" and "occasional" loads and
thermal expansion.
• Electric resistance weld E= LO • Sustained loads. For the 14-in piping in Fig. 3.20, the sustained load
• Double submerged are weld E= 1.0 is pressure, and weight.
• Occasional loads. For simplicity, in this example, we assume that
For example, the 14-in process plant gas piping shown in Fig. 3.20 is there are no occasional loads; pressure transients, high wind, and
fabricated from ASTM A 312 Type 304L stainless steel. It has a design earthquake are not credible events.
pressure of 370 psi anda design temperature of 850°F (the insulation
has been removed in this photograph for inspection purposes). The • Expansion loads. The thermal expansion load corresponds to the
allowable stress for the material at 850°F is S = 12,800 psi. expansion ofthe line as it goes from the ambient shutdown condition
The minimum required wall thickness for the straight sections, to the hot operating condition at 850°F.
according to ASME B31.3 is For the sustained load analysis, the piping system is modeled and
370 X 14 = _ in
the pressure and temperature are entered into the model. The qualifi-
0 20
tmin ~ 2(12, 800 X 1.0 + 370 X 0.4) cation requirement used to evaluate the stress output is given in the
ASME B31.3 code. For sustained loads the longitudinal stresses have
The fabrication undertolerance permitted in the ASTM A 312 mate- to comply with the equation
rial specification is 12.5 percent, so the pipe must be ordered at least
12.5 percent thicker than the minimum required, or PD +0.75i M < kS
4t z
t = tmin X 1.125 = 0.20 in X 1.125 = 0.225 in.
where M = resultant moment dueto sustained or occasionalloads, in·lb
The closest, larger, commercial pipe schedule for 14-in pipe, in accor- Z = pipe section modulus, in3
dance with ANSI B36.19, is schedule 20 with a wall thickness of0.312 in. k = factor, 1.0 for normal operation and 1.2 for occasionalloads
The corrosion allowance for the stainless steel pipe was setas 0.0 in; i = stress intensification factor, given in ASME B31
no wall thinning was expected to occur during service. Otherwise, a S = allowable stress provided in the design code, in this case
future corrosion allowance FCA would have to be estimated over the ASME B31.3, psi
In this example, the computer stress analysis indicates that the high- The pipe stresses are therefore qualified for pressure design (wall
est stress is at an elbow where the stress intensification factor is i = 3.3 thickness and fitting ratings), sustained stresses, and expansion
and the moment is 95,000 in-lb. GivenP = 370 psi, D = 14", t - 0.312", stresses. Th complete the analysis, the designer would also check the
Z - 45 in3, and an allowable stress S = 12,800 psi, the total stress is following:
therefore
• The nozzle loads on the vessels at the two ends of the pipe
370 X 14 + 0.75 X 3.3 X 9 S,OOO = 9376 < 1 X 12,800 • The displacements of the line to make sure they are reasonable and
4 x 0.312 45 that the pipe does not sag, overcome its slope, or interfere with oth-
er components
For the expansion analysis, the longitudinal stress has to comply
with the equation • The design of the supports and their attachments to the building,
ground, or structure
i ~e < f(1.25Se + 0.25Sh)
3.28 Applied Forces
where Me= resultant moment range dueto expansion loads, in·lb Finally, we note that the design equations limit the tensile stress due
f = high cycle penalty factor; f = 1 if there are less than to pressure, bending moments, and torsion; they do not address stress-
7000 cycles of heatup-cooldown es due to applied forces, whether axial tensile forces, compressive
S e = material allowable stress at cold temperature, psi forces, or shear forces. The reason for accounting for applied moments
S, = material allowable stress at hot temperature, psi but ignoring applied forces is that a welllaid-out system, subject to
anticipated design loads does not see large tensile or shear forces.
Note that the expansion stress is checked independently from the Unanticipated loads can, however, cause large shear forces that could
sustained stresses. We have seen earlier that a design life had to be rupture the pipe, typically at its rigid anchor or nozzle points. This is
established to determine the corrosion allowance. Now, with the intro- the case illustrated in Fig. 3.21, where a large accidental impact on the
duction of the factor f we see the need to establish a number of oper- pipe caused it to shear at its nozzle to the equipment, and in Fig. 3.22.
ating cycles over the design life. Ifwe assume one cycle oftemperature
variation from shutdown to hot operation per week, for 20 years, then
the number of cycles over the design life of 20 years is
N - 20 years X 52 weeks/year X 1 cycle/week = 1040 cycles
Because N < 7000, then f= 1, and the allowable stress becomes
Sa 1 (1.25 X 16,700 + 0.25 X 12,800) = 24,075 psi
where S e 16,700 psi is the material allowable stress at ambient 70°F
(shutdown) and sh = 12,800 psi is the material allowable stress at
850°F operating temperature, from ASME B31 or ASME Il.
The computerized expansion stress analysis indicates that the max-
imum expansion stress is at an elbow, with a moment range of 150,000
in-lb, therefore the expansion stress is
150 000
3.3 X • = 11,000 < 24,075 Figure 3.21 Shear rupture at pipe nozzle by accidental impact.
45
3.30 Gas Pipelines

The minimum wall thickness of liquid gas pipelines, transmission or
distribution, is specified in ASME B31.8 and is based on an allowable
stress equal to a fraction F x T of yield13
PD
t=-------
2(FxExTxSy)
where tmin = minimum wall thickness required by ASME B31.4 code, in

P = design pressure, psi
E = longitudinal or spiral weld joint efficiency factor
Sy = specified minimum yield stress (SMYS), psi
Figure 3.22 Crack in welded attachment.11 F is a location class factor

• Location class 1, wasteland, desert, mountains, F = 0.72 or 0.8
3.29 Liquid Pipelines-ASME 831.4 • Location class 2, 10 to 46 buildings within 1 mile, industrial area,
F = 0.6
The minimum wall thickness ofliquid hydrocarbon pipelines, crude oil
or refined products, is specified in ASME B31.4 and is based on an • Location class 3, suburbs, F = 0.5
allowable stress of 72 percent yield, 12 • Location class 4, city, river crossing, F - 0.4
t.
PD
= - -- - - -
Tis a high-temperature correction factor, with T = 1 if the gas is
mm 2x0.72xSy xE below 250°F, and decreasing to 0.867 at 450°F.
For example, consider an electric resistant welded (E = 1.0), 20-in
where lmin-= minimum wall thickness required by ASME B31.4 code, in gas pipeline, API 5L X40 material (specified minimum yield stress
P = design pressure, psi SMYS = 40,000 psi ), buried underground ata ri ver crossing (F = 0.4)
D = outside diameter, in and operating at 1200 psi and 50°F. The minimum wall thickness
E = longitudinal or spiral weld joint efficiency factor required by ASME B31.4 would be
S.1• = specified minimum yield stress (SMYS), psi
t _ = 1200 X 20 =0.75 in
For example, consider an electric resistant welded (ER 1.0), 20-in mm 2 X 0.4 X 1.0 X 1.0 X 40,000
liquid pipeline, API 5L X40 material (specified minimum yield stress
SMYS 40,000 psi), buried underground and operating at 1200 psi At 1200 psi, a 20-in X 0.75-in gas pipeline, API 5L X40, would oper-
and 50°F. The minimum wall thickness required by ASME B31.4 ate at a hoop stress equal to 40 percent of its yield stress
would be
3.31 Fatigue
t . = 1200 x 20 = 0 _42 in
mm 2 X 0.72 X 40,000 xl.O Fatigue failure is a five-stage process. 1
At 1200 psi, a 20-in = 0.42 in-liquid pipeline, API 5L X40, would • Stage 1, microscopic slip bands occur along the planes of maximum
operate at a hoop stress equal to 72 percent of its yield stress. shear stress, roughly at 45 degrees from the applied tensile load.
Figure 3.23 Crack at toe of weld.

Figure 3.24 Pressure fatigue crack at branch weld. 14
• Stage 2, microscopic cracks form alonga slip plane at grain bound-

aries. smooth specimen of the same steel. This effect is even more pro-
• Stage 3, the microcracks evolve into a crack that can be detected by nounced with high-strength steel.
nondestructive examination (NDE). • Overloads. Because the endurance limit is caused by the pinning of
• Stage 4, visible cracks appear and the component will leak when the dislocations an overload could unpin the dislocation and continue
crack progresses through the wall (Fig. 3.23). the crack-opening process.
• Stage 5, the final stage, as the leaking crack keeps growing, the • Residual stress. Compressive residual stresses on the surface tend to
remaining ligament of metal becomes too weak to resist the applied prevent crack initiation, whereas tensile residual stresses can cause
tensile load and fractures. stress corrosion cracking or accelerate the progression of fatigue
cracks.
This last stage will only occur if (a) the leak from Stage 4 is not
• Material. Differences in strength result in differences in fatigue
detected on time, or (b) the material has low toughness and cannot
properties.
hold the crack stable, and (c) there is sufficient force applied to the
crack to open the fracture. • Corrosion. Fatigue cracks open fresh metal to the fluid or atmo-
In practice, the duration of each of the five stages of fatigue depends sphere, causing premature failure.
on severa! factors 1•
There are five methods for the prediction offatigue life oftanks, ves-
• Magnitude of the applied cyclic stress. If the applied stress range is sels, and piping.
below a threshold value, the endurance limit of the metal, then the
crack will not propagate. • ASME Boiler and Pressure Vessel code method. The method applies
to any type of loading, but it is a design technique that includes a
• Existence ofgeometric discontinuities or notches. Discontinuities and safety factor of 2 on the applied stress and 20 on the cycles to fail-
notches represent stress concentrations (Fig. 3.24). ure. It is based on tests in air and may overpredict fatigue life in a
• Preexistence of cracks. If the component has macrocracks introduced corrosive environment.
during construction, then Stages 1, 2, and 3 are nonexistent. • Markl method. The method applies to pipe and pipe fittings under
• Surface finish. Cracks initiate more readily on a rough surface. The cyclic bending or torsional rnoments. It is based on tests in air and
fatigue life of a forged specimen of mild steel is nearly half that of a may overpredict fatigue life in a corrosive environment.
• Fracture mechanics method. The method predicts the remammg In our example, the maximum stress intensity occurs at a carbon steel
fatigue life of a preexisting crack. It applies to corrosive or noncor- pipe-vessel nozzle weld with a fillet cover. The value of S at the
rosive environments. pipe-vessel nozzle weld is calculated for each operating mode, and is
• AWS-AASHTO method. The method applies to as-weldedjoints. It is summarized in Table 3.4. The stress concentration at the local disconti-
based on tests in air and may overpredict fatigue life in a corrosive nuity includes the geometric discontinuity of the weld, but does not
environment. include local flaws such as a weld undercut. These weld flaws are limit-
ed by the workmanship standards of the ASME vessel construction code.
• Fatigue testing. Testing applies to base metal and welded joints, or For tbe purpose of fatigue design, we must defme the design oper-
to full-scale components and equipment. It can be conducted in air ating cycles. In this example, tbe vessel cycles between tbe four oper-
or in a corrosive environment. ating modes as indicated in Table 3.5.
As tbe temperature and pressure vary among modes 1-2, 1-3, 1-4,
3.32 The ASME Boiler and Pressure Vessel 2-3, 2-4, 3-4 (e.g., as the system heats up from condition 1 to 2) so
Code Fatigue Method does the stress at the pipe-to-nozzlejunction, for example, from sl = o
to 8 2 = 40 ksi. In other words, the stress varies through a range as the
The fatigue analysis method adopted by the ASME BPV code is pressure and temperature vary in service. The stress ranges 1'lS are
explained through an example. A pressure vessel in a chemical plant listed in Table 3.6.
operates at the modes presented in Table 3.3. The vessel is made of For each stress range, a stress amplitude (the alternating stress) is
carbon steel with an ASME Code allowable stress Sm = 20 ksi. defined as half the stress range multiplied by a factor Ke
The next step is to calculate the maximum stress intensity at the worst
location for each operating mode. The stress intensity is the largest dif-
ference ofthe three principal stresses Sh 82, and Sa ata point
where S = ASME defined stress intensity, ksi, and S; = principal stress

TABLE3.4 Stress lntensity for Each Operating Mode
in direction i, ksi, i = 1, 2, 3.
Mode Pressure, psi Temperature, oF Stress S, ksi
The calculation of the principal stress must be based on an accurate
1 - Atmospheric o 80 o
model of the vessel, including stress risers at local discontinuities. For
example, the stress S at a fillet weld could be three to four times the 2 -Normal operation 500 400 40
nominal stress away from the weld discontinuity. Alternatively, to 3 - Hot operation 800 500 50
obtain S, the nominal stress away from the discontinuity is multiplied 4 - Feed operation 200 250 20
by a fatigue strength reduction factor (FSRF) of 3 to 4 to obtain the
stress at the discontinuity.
TABLE 3.5 Design Operating Cycles
Range n, cycles
TABLE 3.3 Vessel Operating Modes
1-2 10,000
Mo de Pressure, psi Temperature, °F 1-3 1,000
1 - Atmospheric o 80 1-4 1,000
2 -Normal operation 500 400 2-3 10,000
3 - Hot operation 800 500 2-4 1,000
4 - Feed operation 200 250 3-4 1,000
TABLE 3.6 Deslgn Operating Cycles TABLE 3.7 Fatigue Usage Factor
Range 11, cycles J.S, ksi ASMEN, Actual n,

Range J.S, ksi Sslt. ksi cycles cycles n/N
1-2 10,000 40
1- 2 40 20 100,000 10,000 0.100
1-3 1,000 50
1-3 50 25 42.500 1,000 0.024
1-4 1.000 20
1-4 20 10 > 106 1,000 0.0
2-3 10,000 10
2-3 10 5 > 106 10.000 0.0
2-4 1,000 20
2-4 20 10 > 106 1,000 0.0
3-4 1.000 30
3-4 30 15 350,000 1,000 0.003
Total 0.127
where 8 811 = alternating stress between two conditions of operation, ksi

K. elastic-plastic stress correction factor
t:.S stress range between two conditions of operation, ksi
The stress correction factor K,. corrects the elastically calculated
stress to account for plasticity, if any. The factor K, depends on the
magnitude of the primary plus secondary stress intensity (SP+S) com-
pared to the ASME code material allowable stress Sm.
For Sp s < 3 S,.,
K., = 1
For 3 Sm < Sp.,r; < 3 mS,.,
K
"
=1+ 1 -n
n(l-m)
(~-1)
3Sm
o '
For Sr s > 3 mS,., 1e+4 +---t--t-t-t-trn+-

1e+1 1e+2 1e+3 1e+4 1e+5 1e+6
N umber of cycles
Figure 3.25 General form of S·N ASME fatigue curve for carbon steeJ.l6
where Sm - code allowable stress and m and n = material constants.
For low-alloy steel m = 2 and n = 0.2; for carbon steel, m = 3 and n =
0.2; for austenitic stainless steel m = 1.7 and n = 0.3. of permitted design cycles N is obtained from the ASME fatigue curve
Because the highest stress range intensity does not exceed 3Sm = 3 X for the material (Fig. 3.25 or Table F.9 of API 579).15 The number of
20 ksi, then there is no need for a plasticity correction factor and there- actual operating cycles n for a given range is obtained from the design
fore K.,. = 1. prediction or, in operating systems, from operating records and future
The operating ranges, the corresponding stress ranges t:.S, and the projected operating modes. The fatigue usage factor n/N is calculated
stress amplitudes S. 11 can now be summarized in the first three for each range, and then added to obtain the total usage factor for the
columns of Table 3.7. For a given alternating stress 8 811 the number design life ofthe component. Ifthis total usage factor is less than 1.0,
as is the case in this example where the total fatigue usage factor is where Samplitude = stress amplitude at the point, psi, and N = cycles to
0.127, then the component fabricated in accordance with the ASME failure.
code will not fail by fatigue for the calculated stresses and cycles.
The ASME S-N fatigue curve, such as shown in Fig. 3.25 for carbon For example, if a fillet weld in a carbon steel pipe is subject to a cyclic
steel, is a design curve, established on the basis of cyclic, displacement stress range of 20 ksi (stress amplitude of 10 ksi), then the number of
controlled, fatigue tests on a smooth base metal specimen, tested in air. cycles to fatigue failure of that weld is
Displacement controlled means that the specimen was cycled by stretch-
ing to a fi.xed imposed strain ~ê-..ctual· As the specimen starts to crack, the
stress in the specimen drops but the stretch, the strain, does not change
from ~êactual· For each specimen, the number of cycles to failure N was
plotted against a pseudoalternating stress Salt that was calculated as where i = 2.1 is the maximum stress intensification factor for a fillet
weld, it corresponds to a fillet weld with a short leg, and is obtained
S _E X &actual from ASME B31.3. Therefore
alt- 2
N- 216,000 cycles to failure.
where Salt = pseudoelastically calculated stress, psi
E = Young's modulus ofthe material, psi
3.34 Example of the Markl Method
~êuctuaJ = actual strain range recorded during the test, may be
in Víbration
elastic or plastic
Figure 3.26 shows a pipeline that failed from vortex-induced vibration.
The measured (N, Sa~t) data were then corrected by a factor of 2 on To explain similar failures, consider a pipeline in open windy terrain, or
stress and 20 on cycles to failure. The margin of 20 on cycles corre- a subsea pipeline subject to significant vibration dueto vortex shedding
sponds to the following corrections: as the wind or sea current crosses the pipeline. The bending stress in
• A factor of 2.0 to account for test data scatter the vibrating pipe span can be approximated by the stress at midspan
ofa beam
• A factor of 2.5 to account for the small size of the test specimen (the
smaller the specimen the less chance of material flaws)
• A factor of 4.0 to account for surface finish (the tested specimens
were smooth, which gjves a better fatigue life than a rough surface
finish or weld ripples) and-arguably-for corrosion (the tests were
where = bending stress, psi
CJ&
E = modulus of elasticity, psi
conducted in air) D = pipe diameter, in
L = span length, in
3.33 The Markl Fatigue Method d = midspan deflection, in
In the 1940s and early 1950s, A. R. C. Markl performed a series of
fatigue cyclic stresses on pipe fittings: elbows, tees, mitered bends, and
so on. From these tests, Markl established a simple and elegant rela-
tionship between the imposed cyclic stress amplitude S amplitude and the
number of cycles to failure N.l 7- 19 According to Markl's work, which is
the cornerstone of today's ASME B31 design equations, fatigue failure
(through-wallleakage) is predicted to occur when
. 245,000
iSamplitude = N0.2
Figure 3.26 Vortex shedding induced fatigue. 14
T
In the case of a 20-in carbon steel line pipe, on 30-ft-long spans,

which deflects 0.45 in at midspan, the bending stress is
6
Gb =6 X 30x 10x20 .
X 0.45 =12,500 psl = 12.5 ks1
.
(30x12)2
At the stress amplitude of 12.5 ksi, the number of cycles from the
ASME design curve (Fig. 3.25) is 1,000,000. Because the ASME da
fatigue curve includes a safety factor of 20 on cycles, the actual dN
cycles to failure are approximately 20 X 1,000,000 = 20 million
cycles to failure. If the vibration has a frequency of 1 cycle/2 seconds
= 0.5 Hz, it would only take 463 days to fail the pipe ifthe vibration
were continuous.
20x106 1
t= x = 463 days
0.5 86,400
.1-K
3.35 The Fracture Mechanics Figure 3.27 Crack growth ratc zones.
Fatigue Method
Fracture mechanics and the calculation of the stress intensity at a
crack tip K is the subject of Chap. 9. The fracture mechanics method • Type of material
of estimating the remaining life of a crack under cyclic load is pre- • Environment <fluid, temperature) around the crack
sented here rather than in Chap. 9 in order to group all the fatigue • Ratio R = Kmax1Km.n of the rnaximum and minimum stress intensi-
methods in a single place. The fracture mechanics approach predicts ties driving the crack
the progress of an existing crack. As a fatigue crack of size a propa-
• Loading rate (speed)
gates, the relationship between the increment in crack size da and the
range of applied stress intensity M is given by For example, a branch connection, such as shown in Fig. 3.28, has a
crack subject to pressure fluctuations that cause an alternating applied
stress intensity K at the crack tip. The calculation of K is presented in
Chap. 9. For the purpose of this example, assume that K is calculated
where a = crack size, in to fluctuate between + 20 ksifu and O ksifu every 10 seconds. If the
da crack growth, in component operates in a marine environment, how long will it take for
N number of cycles the crack to progress 0.1 in (2.5 mm)? Why will the actual crack grow
dN number of incrementai cycles_ even faster than calculated?
t::J:( - range of stress intensity, ksi,'in For steel in a marine environment up to 54°F (20°C) the relationship
m, n - parameters between crack growth rate da/dN and applied stress intensity range
t::J:( isl5
This relationship is represented by zone II on the curve in Fig. 3.27.
Zone li is preceded by a crack formation Zone I, with a threshold ÃK1h, :~ = 3.80 X 10-9 X (t:J{)"
and followed by fast rupture Zone III.
The parameters m and n that govern the crack growth rate depend
on severa] factors.2o : in / cycle - 3.80x10-9 x (20 ksi~)
r-------------------~==========~1000
-+- Class 124+
···<>··· Class 100
-+ - Class 80
8!. 1000 -ll'····Ciass 63 'iii
:2 100 .X
C/)~
u) -- Class 50
Q)
O)
-·o-· Class 40 Q)
O)
c c
~ ~
(/)
~ 100 (/)
~ ~
êi) êi)
10
1e+2 1e+3 1e+4 1e+5 1e+6 1e+7 1e+8 1e+9
Number of cycles N
Figure 3.29 Fatigue curveA for welds.15.21.22
base metal specimen of the ASME Boi ler and Pressure Vessel Code.
Figure 3.28 Pressure fatigue at branch.11
The rules were improved through the 1960s and 1970s accounting
for the stress range crmax to crmin and a statisticallower bound with
confidence levei of 95 percent survival at 95 percent confidence.
da 9 1 cycle . O1 .
a =-N = 3.80x10- x20x---xtJme= . m Each weld detail, for example, a butt weld or a fillet-welded attach-
dN 10 s ment, is assigned a class and each class is assigned a fatigue curve.
Time= 0.1 in x 10 sx 109 x 1 -150 days A similar approach is applied to pr essure vessels in British
3.80x20 3600s/hx 24h/day Standard 5500.22
In Figure 3.29 the weld is classified on the basis ofthe following.
The actual crack will grow to 0.1 in sooner than 150 days because,
as the crack size a increases so does the stress intensity K and there- • The direction of the fluctuatíng principal stress relative to the weld
fore !J.K and da/dN {rate of crack propagation). Also, this assessment detail
does not accounL for residual stresses introduced during mill fabrica- • The location of possible crack initiation at the weld detail
tion of the line pipe by cold bending a skelp Cplate) ínto the pipe cylin-
drical shape, or the weld residual stresses introduced when welding • The geometrical atTangement and proportions of the weld detail
the branch to the header. • The methods of manufacture and inspection
For example, a vessel is subject to a fluctuating pressure hoop stress

3.36 The AWS- AASHTO Fatigue Method range crr = 10 ksi; the longitudinal weld in the vessel is a class 80
In the 1940s the American Welding Society (AWS) developed fatigue detail (Fig. 3.30). From Fig. 3.29 its fatigue life is 3 million cycles.
life curves for railroad bridge weld details. The fatigue curves were
adopted by the American Association of State Highway Officials 3.37 Fatigue Testing
(AASHTOl in the forro of cycles to failures as a function of maxi-
mum cyclic stress crmax and the ratio of crm;nlcrmax·21 The design life A specimen of base metal or weld, or a full-scale component or equip-
was based on tests of actual weldments, as opposed to the smooth ment can be fatigue tested, typically by imposing a controlled cyclic
11 O Chapter Three Design 111
• Classify the stresses along the cross sections as membrane, bending,

and peak.
• Compare membrane, bending, and peak to ASME code limits.
Figure 3.30 Class 80 weld and cyclic load. The concept of stress classification is used in ASME VIII Div.2 and
ASME III Div.l, and is explained in Fig. 3.32.
Loads such as weight, pressure, temperature, and externai forces
are applied to a fine-mesh elastic finite element model that closely
models materiais, shapes including discontinuities such as welds, and
boundary conditions. Stresses are obtained at each point ofthe model.
'Ib evaluate these elastic stresses in accordance with ASME VIII Div.2
and ASME III Div.l the stresses have to be classified. At every cross
section in the model the stress distribution through the wall, as shown
in Fig. 3.32 at left, is written as the sum of a constant (membrane)
stress A, = B,., a linear (bending) stress through zero at midwall Bb = O,
and the balance (peak) stress such that
The average stress along this line, the membrane stress, is
where <Jm = ASME membrane stress in a cross section, psi

crL = total stress at every point along cross-section integration
line, psi
Figure 3.31 23
Full-scale fatigue testing ofvessel supports. t wall thickness or, more generally, thickness of line of
=
stress classification, in
x = distance along line of stress classification, in
displacement, as did Markl in his tests on pipe fittings, or by appl~ng The maximum bending stress along the line of stress classification is
cyclic forces or moments.23-27 Figure 3.31 illustrates a full-scale fatigue
test of a vessel on supports.
A
3.38 ASME Stress Classification along a Li ne
The design-by-analysis rules of pressure vessels in accordance with
ASME VIII Div.2 and ASME III Div.l consist of six steps. ,___~_
B_ __ ____ ,....__~
• Prepare a finite element elastic model of the component.

• Apply the loads.
• Calculate elastic stresses at ali points of the model.
• Select cross-sectionallines through the component thickness. Figure 3.32 Stress classification.
112 Chapter Three Design 113
where crb = ASME maximum bending stress in a cross section, psi. where PE = eiastic collapse externai pressure, psi
t = pipe wali thickness, in
The balance between the total stress crL at each point and the cross D = pipe outer diameter, in
section's membrane stress crm plus the maximum bending stress crb, the E = Young's modulus, psi
peak stress, is v = Poisson ratio
For exampie, the eiastic collapse externai pressure for a tank with a
diameter D = 30ft = 360 in, and thickness t = 0.50 in,
When instead of a fine mesh through-thickness, the stresses are 3
6
only known on the inner and outer surface of shell elements, then the p =2x30x10 (0.5) , 0 _2 si
membrane stress in the shell element is E 1-0.32 360 p
Not much externai pressure is needed to buckle Iarge thin tanks. This
is illustrated in Fig. 3.33. In this case, the vacuum breaker on the tank
top was covered by a plastic sheet placed during maintenance painting.
where crin = stress at inner surface of shell, psi, and crout = stress at When liquid was pumped out ofthe tank, the plastic sheet was partial-
outer surface of shell, psi. Iy sucked into and plugged the vacuum breaker. Pumping continued,
drawing a negative pressure inside the tank, causing it to buckle. In Fig.
The bending stress in the shell element is 3.34 a rail tank was cleaned with steam and then sealed. The trapped
steam cooled and condensed, drawing a negative pressure inside the
tank, causing collapse under the externai atmospheric pressure.
For small Dlt (thick cylinder) buckling results from yielding of the
cross section. Yielding occurs at a pressure Py, given by
For fatigue analysis, the peak stress is obtained using a fatigue
strength reduction factor
aF = (<Tm +abXFSRF-1).
In addition to the stress decomposition into membrane, bending,

and peak, stresses are also classified as primaiy, secondary, and local
in accordance with ruies provided in ASME VIII Div.2 or ASME III
Div.l. Once stresses have been classified, the proof of design is
achieved when each category of stress and stress combination is shown
to be within a code-specified allowable stress.
3.39 Externai Pressure

The evaluation offitness-for-service for vacuum systems and external-
ly pressurized systems requires the understanding of the rules for
externai pressure design. For Iarge Dlt (thin wall cylinder) buckling
under externai pressure occurs whiie the material is still elastic (elas-
tic buckling). The elastitc collapse externai pressure is PE given byt,29-33 Figure 3.33 Collapse of storage tank under externai pressure.
Design 115
11 4 Chapter Three
6 ASTM A 182 Standard Specifu:ation for Forgcd or Rolled Alloy anel Stainles.~ Steel
Pipe Flangcs, Forged Fittings, and Volves and Parts for lligh-Temp<'rature Service,
ASTM Intcrnational, West Conshohocken, PA.
7. ASME Bl6.11. Socket Welding and Threaded Forged Steel Fittings, American
Society of Mechanical Engineen;, New York.
8. Catastrophic Vessel Ot•erpressurizaton, U.S. Chemtcal Safety and Hazard
Investigation Board Report, Repor t No. 1998-02-1-LA, Washington, DC, 1998.
9. ASME Boiler and Pre~sure VesAel Code, Section VIII, American Society of
Mechanical Engineers, New York.
10. ASME B31.3. Process Pípíng, American Society of Mechanical Engineers, New York.
11. Thielsch Engineering, Cranston, RJ.
12. ASME B31.4, Liquid Petroleum Transportation Piping, American Society of
Mechanícal Engineers, New York.
13. AS~1E B31.8, Gas 'IransmissLOll and Distributwn Piping, American Society of
ME>chanical Engineers, New York.
14. Kiefner & Associates, Worthington, OH.
15. API RP 579, Fitness-for-Service, American Petroleum Institute, Washington, DC.
16. ASME Boiler and Pre!<sure Vessel Code, Section ll, Materiais, American Society of
17 Markl, A. R. C., Fatigue tests of we Idi ng elbows and comparable double-rniter bends,
Transactions o{ the ASME, 69(8), 1947.
18. Markl, A. R. C., Fatigue tests of piping components, Transactwn.~ o{ the ASME.
74(3,, 1952.
Figure 3.34 Collapse of rail tank under externai pressure. 19. Markl. A. R. C., Piping Oexibility analysis, 'Iransaction.~ o{ the ASME, February,
1955.
20. Bannantine, J. A., et. ai., Fundamentais o{ Metal Fatigue Analysis, Prentice· Hall,
Englewood Cliffs, NJ
21. Bnrsom, .J. M., Vecchio, R. S., WRC Bulletin 422, Fatigue of Welded Structures,
t Pressure Vessel Research Council, 1997.
2-SYI
D I 22. Enquiry Case BS5500/79 May 1988, Assessment of Vessels Subject to Fatigue:
Alternatil'e Approach to Methocl in Appendix C, British Standards lnstitute,
England. British Stanclard Specification for Unfired Fusíon \Ve/ded Pressure VcRsels,
where Py = extemal pressure at yielding, psi, and S} • = minimum mate- BS5500, Issue 1. British Standards Institute, England, 1991.
rial yield strength in the hoop direction, psi. 23. ASTM E 466, Standard Practice for Condu.cting Force Controlled Constant
Amplitude Aâal Fatigue Tests o{ Metallic Materiais, ASTM International, West
Conshohocken, PA.
At intermediate values of Dlt the buckling regime transitions from 24. ASTM E 468, Standard Practice for Presentatton ofConstant Amplttude Fatrgue Test
elastic collapse PE to yield Py, with a collapse pressure R(•sults for Metallic Matrr~ais, ASTM Intemational, West Conshohocken, PA.
25. ASTM E 606, Standard Practice for Strain-Control/ed Fatigue Testing, ASTM
P. - PyPE lnternational, West Conshohocken, PA.
c- f 2 2
26 ASTM E 647, Standard Thst Metlwd for Measurement of Fatigue Crack Grou·th
"Pl'+Pe Rates, ASTM lntemational, West Conshohocken, PA.
27. ASTM E 1823, Standard Thrminofo!(y Re/ating to Fatigue and Fracture 'll>,çfíng,
ASTM I niernational, West Conshohocken, PA.
where Pc = collapse pressure, psi. 28. Paulin Research Group, Houston
29. ASME VIII, ASME Boiler and Pressure Vessel COO.e. Section VIII. Division 1. Rufes
for Constructton of Prc.~sure Vessels. AG-28 Thickness o{ Shells and ThbeH wzder
References Externa! Prcssure, American Society of Mechanical Engineers, Ncw York.
30. Bednar, H. H., Pressure Vessel De.çign Handbook, Kricgcr, Melbourne, FL.
1. Antaki, G. A., Piping and Pipeline Engineering. Dekke~·· _New York. . 31. Den Hartog, Advanced Strength of Materiais, Dover Publications, New York.
2. ASTM A 335, Standard Specification for Seamless Femt1c Alloy-Steel Ptpe for Htgh- 32 Farr, J. R., Jnwad, M. H., Guidebook for the Design of ASME Sectron VIII Pressure
Temperature Seruice, ASTM lntemational, West Consho~ocken, ~A. Vessels, ASME Press, New York.
3. ASME B31.1 Power Pipil1g, American Society of Mechamcal Ipngmeers, _New Yo~k. 33. Harvey, J. F.. Theory and Design of Pressure Vessels, Vnn Nostrand Reinhold, New
4. ASME Bl6.9, Factory-Made Wrought Steel Butt Welding Fittmgs, Amencan Soctety York.
ofMechanical Engineers, New York .
5 ASME Bl6.5, Pipe Flanges and F/anged Fittings. NPS ~ through NPS 24, Amencan
Society of Mechanícal Engineers, New York.
Chapter
4
Fabrication
4.1 Fabrication and Construction Flaws

For clarity, we use the term fabrication to refer to the operations tak-
ing place in the mill or shop, and construction to refer to field erection.
All tanks, vessels, and piping contain, to varying degrees, fabrication
and construction flaws in base metal and joints. These flaws exist
because of the following:
• The flaws are detected but are smaller than permitted by construc-
tion codes, in which case they are called discontinuities rather than
flaw~. and left as-is.
• The flaws have eluded initial examination and testing during fabrica-
tion and construction. This is possible because: (1) in most cases the
material standards and construction codes do not require full volu-
metric examination ofbase metal, or {2) the construction codes do not
require 100 percent volumetric examination of welded joints (with
some exceptions such as ASME VIII lethal service vessels and ASME
B31.3 high-pressure piping, and oil and gas pipelines in critica]
areas), or (3) the initial nondestructive examination missed the defect.
Fabrication and construction flaws are important to fitness-for-service
assessment for two reasons:
• They can grow in service or aggravate a service-induced flaw.
• They can be detected during in-service inspections and be confused
with a degradation-induced flaw.
117
118 Chapter Four Fabrlcatlon 119
4.2 Base Metal Defects the ingot is rolled into plates or strips. They can also resu!t from
Defects can be introduced into the molten metal or appear during forg- large ínclusions.
ing, rolling, extrusion, or casting: • Scabs are defects often dueto scale rolled onto the surface (Fig. 4.3).
• Laminations in plates are metal separations parallel to the surface • Gouges are knifelike cuts of the material surface (Fig. 4.4).
(Figs. 4.1 and 4.2). They come from voids formed when the molten
metal solidifies in the mill, and these voids are then elongated when
Figure 4. 1 Plate failed at lamination. 1

Figure 4.3 Scabs on inner surface.l
Figure 4.2 Laminations in plate. Figure 4.4 Gouge on metal surface.'

Figure 4.8 Weld offset in spiral-

welded line pipe.
Figure 4.5 Lap evolved into through-wall crack.1
Flaws can be introduced in each of these steps; this is why material

specificatíons have explicít limits on the type of flaws that are accept-
able and when flaws can be repaired. It is the duty of the fab1icator,
and the responsibility of the owner's ínspector to ensure that these
material and code specification limits are met. An example of a mill
fabrication defect is the eccenbic weld in Fig. 4.8, where the two ends
of the plates were not correctly aligned before welding.
4.4 Welding Techniques

There are many excellent publications on the technology and art of
welding. 4 12 From a fitness-for-service perspective, it is worthwhile to
summatize some salient features of weldin g that will affect component
integrity. Are welding is the common weldíng technique used in the
fabrication of tanks, vessels, and piping. The broad principies of are
welding are
Figure 4.6 Grinding surface lap.3 Figure 4.7 Crack initiating at oxide
inclusion.2 • An are is formed between the part and an electrode.
• The are temperature melts the part to be welded, forming a weld
• Laps are folds in the metal, forced flat during plate r olling (Fígs. 4 .5 pool.
and 4.6). • Filler metal may be added or the part may be welded without filler
• Inclusions are trapped impurities (Fig. 4. 7). metal (autogenous welding).
• Gas or flux protects the weld pool from the environment until it has
4.3 Fabrication Flaws
solidified.
• The weld procedure is qualified, and the weld specification defines
Shop fabrication of vessel heads and components a~d mill fabrica~ion
the essential variables.la.J4
of piping involve rolling, forming, extruding, weldmg, and bendmg.
122 Chapter Four
Fabrlcatlon 123
• The welder is trained and qual:ified to the weld procedure. 13,14

• Welding may be manual or automated.
The common are welding techniques for tanks, vessels, and piping are
• Shielded metal are welding (SMAW)
• Flux core are welding <FCAW)
• Gas tungsten are welding (GTAW, TIG)
• Gas metal are welding <GMAW, MIG, MAG)
• Submerged are welding (SAW)
Shielded metal are welding is also referred to as manual metal are
welding or stick welding (Fig. 4.9). The consumable eleetrode is a Figure 4.10 Flux core are weld-
ing.
stick (rod} protected by a metallie sheath (flux), hence the name
shielded metal.
The flux core are welding eonsumable eleetrode eontaills in its center
either a flux that automatically shields the molten pool (self-shielding
FCAW), or minerais or alloys, in which case gas shielding is necessary
(gas-shielded FCAW; Fig. 4.10).
Gas tungsten are welding is also referred to as tungsten inert gas
welding. The nonconsumable electrode is made of pure tungsten, zir-
conium tungsten, or thoriated tungsten. The eleetrode is shielded by
welding-grade argon or helium flowing through the same nozzle.
Welding may be autogenous (no filler metal) or a separate welding
rod may be used to supply filler metal, as illustrated in Fig. 4.11. The
are and weld pool are protected by inert gas (welding grade argon or
helium).
Figure 4.11 TIG welding.
Gas m~tal are weldi_ng is also referred to as metal inert gas welding
whe~ shielded b~ an mert gas such as helium, and metal active gas
weldmg when sh1elded by a reactive gas such as C02• The consumable
electrode is fed t~rough a nozzle that also supplies a shielding gas
such as argon, hehum, or carbon dioxide.
Figure 4.9 Stick Welding. In submerged are welding granular flux is continuously poured over
the are and molten metal to shield them from the atmosphere (Figs.
4.12 and 4.13}.
Fabrication 125
124 Chap ter Four
V Mo Mn+Cu+Cr Si Ni
CE =5B + C . +-+ +-+-
10 15 20 30 60
For a carbon content above 0. 1 2~ the carbon equivalent is defined as
Cr+ Mo+ V Mn Ni+Cu

CE =C + +-+---
5 6 15
For ease of weldability of carbon steel, CE is usually limited to 0.3 to
0.4 maximum, depending on the type of steel and welding technique.
4.6 Weld Quality

Weld quality is achieved through a combination of three elements that
ar e explicitely addressed in construction codes such as ASME IX and
API 1104:13,14
• Weld procedure
Fig ure 4.12 Submerged are welding. 16 • Welder qualification
• Weld quality control
Welding procedure. A qualified welding procedure that defines criticai
welding parameters, including weldjoint design, welding position, fix-
ture, weld backing, composition of filler metal and flux, type of elec-
trode, electrode diameter, W€lding current, electrode-work gap (Fig.
4.14 indicates how the gap can be controlled by using a weld rod),
travei speed, welding technique, voltage, shielding gas flow, preheat,
interpass temperature control, and postweld heat treatment.
Figure 4.13 Submerged are welding. 15
4.5 Carbon Equivalent

An important parameter in the weldability of steel is its car~on_ e'!uiv-
alent. Generally, the lower the carbon equivalent, the eas1er 1t 1s to
weld. For example, the carbon equivalent for a steel with a carbon con-
Figure 4.14 Weld preparation root openíng.
tent at or below 0.12o/c is16
126 Chapter Four Fabricatlon 127
Figure 4.1!1 Wehl qualification bend specimen.

Figure 4.16 Branch connection with welder's identification.
Welder qualification. First, a welder must undergo a period of

apprenticeship. Then , the welder must be qualified. Qualification is • Overheating the fluid to the point where it is a chemical or explosive
achieved by welding test samples, following a specific weld proce- hazard
dure, and the weld samples are examined and tested for defects. • Burning through the wall (blowout)
Figure 4.15 is an example ofthe weld bend test. This particular test • Causing weld defects, particularly cracks
failed as the specimen developed a crack during the bend test. A suc-
cessful welder is qualified to weld a group of similar materiais. The Preventlng hazardous conditions. There ar e two forms ofhazards when
welder's qualification has to be maintained up to date. welding online.
Weld quality control. Construction codes define weld quality control
requirements in the following terms. • Overpressure. If isolated and trapped, the heated fluid will over-
• How the welds have to be examined (visual, surface PT or MT, pressure, possibly causing a rupture.
volumetric RT or UT) • Toxicity, flammability, or explosive potential. Some process media
• How many welds have to be examined (a percentage or all) simply should not be heated, even when flowing, because they could
• The acceptance and rejection criteria for weld indications cause a toxíc release or create the risk of an explosion. In this case,
• The progressive sampling rat e if a weld fails the line should not be repaired in-service, and may even have to be
cleaned after flushing and before welding. Examples of process
Each weld is identified against the welder, as illustrated in Fig.4.16, media that should be properly flushed from the system prior to
and when a weld fails, welds made by that welder are reassessed. repair include hydrogen, hydrogen cyanide, oxygen, caustic materi-
ais, chlorine, and others.
4.7 Welding in Service
Preventing burn-through (blowout). Burn-through will happen if the
In many cases it ts possible to weld on the component while the system metal below the weld pool is too weak to contain the internai pressure.
is in service.l7.lti But the operation has to be prequalified and planned To prevent this condition, the inner surface metal temperature has to
carefully. The in-service welding procedure has to be qualified to prevent be maintained below 1800°F, or even less for certain materiais and
three adverse outcomes. high pressures. There are severa) ways to achieve this objective.
• Do not weld in-service on wall thickness below Y!s in. Preventio n of cracklng. Because of heat transfer with the flowing flu-
• Experience to date indicates no risk of burn-through when welding id, the weld bead cools faster when welding in-service than when weld-
in service on wall thicker than X in, but a confmnatory check of the ing on an empty component. This can lead to a bard martensitic
strength of the metal at the welding temperature is necessary. structure prone to immediate or delayed cracking, as discussed in
Chapter 2, and illustrated in Figs. 4.18 and 4.19. The concern with
• Use small electrodes 0~2-in electrode for wall thickness between >'•s in hydrogen ditTusion is typical of ali welding operations, but is more crit-
and X in; 1,(-in electrode for wall thickness between X and X in). icai here beca use of the potential for a harder microstructure. To pre-
• Keep the weld current as low as practical to prevent overheating. vent these conditions, consider the following.
• Avoid heating a trapped fluid which would cause a pressure rise. • Preheat the metal.
• Use a computer model simulation ofthe welding process to pred~ct the • Use a low-hydrogen electrode, with strict contrais to prevent hydrogen
temperatW'e of the inner wall, and confirm that it is not excessiVe. contamination.
• Qualify the welding procedure and the welder on a mockup with cir- • Achieve low-carbon equivalent welds.
culating flow (Fig. 4.17).
• Reduce residual stresses where possible by reducing the size of the
• Reduce the system pressure if possible. It is a common practice, weld repair, thereby reducing the extent of weld shrink.age.
when wP-liling in-service pipelines, to reduce the line pressure by at
least 20 percent. • lfpossible, reduce pipeline restraints during welding and cooling; in
ali cases, prevent large pipe movement.
• Immediately postheat the weld zone between 300 and 600°F.
• Reduce cooling rate. Confirm low cooling rate by measurement on a
prototype mockup, with flow.
• For wall thinner than }:z in, minimize flow rate to reduce weld cooling
rate.
• Slow down electrode travei speed.
• Use a temper bead deposition procedure.
• Verify the hardness of completed welds to be below at least 350 HV.
Figure 4.17 In-service welding qualification test.• Figure 4.18 Cracking of intermcdiate weld passes.
130 Chapter Four Fabrlcation 131
Figure 4.19 Through-wall crack

in weld.
Figure 4.20 Gas leak detector.
• Qualify the welder and the welding process on a prototype mockup,

with flow. by bubble solution, pressure drop, specialized spectrometer, or sound
• Perform volumetric examination of the in-service weld, after welding and ultrasound detectors as illustrated in Fig. 4.20.
but also after 72 h to detect any evidence of delayed hydrogen cracking.
4.9 Pressure o r Leak Testing- Why?
4.8 Pressure or Leak Testing- How?
There are many reasons for conducting pressure or leak tests, not the
The last step offabrication or construction typically is pressure or leak least of which is that they are required by construction codes and reg-
testing. There are three general types of pressure or leak tests. ulations. Reasons for pressure or leak test ing include
• Apressure test may be defined as a test ator above the normal oper- • Detection of leaking mechanical joints (flanges, threads, etc.).
ating pressure. In many cases pressure tests are either hydrostatic • Detection of through-wall cracks or defects in welds. For example,
or pneumatic and are conducted at the design pr essure to 1.5 times the pipe in Fig. 4.21 fractured during the hydrostatic test. The frac-
the design pressure. ture was traced back to a naw in a valve weld. In thls case, the test
• An in-service leak test is a test that consists simply of visually uncovered a weld flaw.
inspecting the joints while starting up the system. If a leak does • Detection of through-wall cracks or defects in base metal.
occur it is repaired. This type of test is linúted to the least criticai
service, where a leak during startup, and subsequent cleanup and • Detection of part-wall defects that are burst open during pressure
repairs, would be acceptable. test.
• Sensitive leak tests are tests that can detect a leak of at least 1o-a • Mechanical stress relief, as will be explained in Sec. 4.20.
cdsec. They include the bubble test (ASME B&PV Section V Article • Rounding and blunting of cr ack tips. On the other hand, it can be
10, Appendix I), the vacuum box test (ASME B&PV Section V Article argued that the pressure test may extend the crack length. Under
10, Appendix li), and the helium-sensitive leak test (ASME B&PV the best of circumstances this may cause a leak t h at will therefore
Section V Article 10, Appendix IV and V). 19 Gas leaks can be detected be detected and repaired prior to service, but under the worst of cir-
pipelines the maximum allowable operating pressure is set as a frac-

tion of the test pressure.21
4.10 Pressure or LeakTesting-Cautlons

Hydrostatic testing (hydrotest) should be perfor med with a minimum
quality of test water:22
• Potable or fresh water, treated with at least 0.2 ppm chlorine.

• Chlorides below 50 ppm.
• Odorless (no hydrogen sulfides).
• pH between 6 and 8.3.
• Temperature below 120°F and above 70°F.
• With potable water, exposure not to exceed 21 days.
• With frPRh wflt.er, exposure not to exceed 7 days.
• Drain and wash with potable water.
• Dry.
During hydrostatic testing the system joints are examined for evidence
ofleaks. When permitted by code and regulations, direct visual inspection
for leaks may be replaced by other detection techniques, which include
Figure 4.21 Valve and pipe failure dut;ng hydrotest. 1
• The test pressure once established can be tracked on apressure-time
chart for evidence of a pressure drop indicating a leak.
cumstances the crack will grow during pressure testing, getting • A tracer gas can be added to the hydrotest water; th e gas would seep
worst before operation, without leaking and therefore without being out at a leak site and small amounts can be detected from the surface.
detected. In this case, the pressure test has created a worst condition.
• Coloring can be added to the hydrostatic test water to test fully
• Testing the integdty of the support system under full weight of buried pipe by looking for colored water on the ground surface.
water-filled component.
• Detection of potential for brittle fracture. This however should never A practice followed to make sure that a large pipeline does not yield
be the case, since the potential for brittle fracture should be eliminated during pressure testing is to use a plot of the water volume added ver-
at the design stage. sus line pressureY A deviation of the plot from a straight line indi-
cates that the elastic limit of some of the pipe within the section
• Confirmation of complction of installation; no missing parts or fit-
has been reachcd.
tings.
• Check of design, but not its confirmation, because test pressure (on
the order of 1.5 times the design pressure) is lower than code expected 4.11 Test Pressure forTanks
design margins (on the order of 3 to 4 times the design pressure). Oil storage tanks are filled with water and examined for leaks. The
• As a means of fitncss-for-service assessment. For example, for stor- wall thickness of a tank is calculated based on the fill height and
age tanks "the effectiveness ofthe hydrostatic test in demonstrating hydrostatic design fill height, using the "1 foot method" of API 650.
fitness for continued service is shown by industry experience."20 In The design thickness for hydrostatic testing is as follows.22
stext
2.6D (H -1)
tt=----- ppnnumatic =1.1 X MAWP X -S- -
st design
where t1 - hydrostatic test shell thickness, bottom course, in

D tank diameter, ft 4.13 Test Pressure for Power Piping
H = design liquid levei, ft The hydr ostatic test pressure of power piping is24
S1 = hydrostatic test stress, psi
Phydro = min(l.5 Pd, ,;gn; Pcomp)
where P design = design pressure, psi, and Pcomp = maximum allowable

test pressure of any nonisolated components, such as vessels, pumps,
where Sy material minimum specified yield stress, psi and Su =mini- or valves.
mum specified ultimate strength, psi.
In a ddition,
The permitted hydrostatic test stress S 1 is approximately 10 per cent
P.hydro x D ~ 0 .9 X S
larger than the permitted design stress sd given by
2xt Y
Sd = min(~Sy;~Su) P.
hydro
xD M
+ Q.75i DL+LL s; 0.9 X S
4xt Z Y
4.12 Test Pressure for Pressure Vessels wher e D = pipe diameter, in

t = wall thickness, in
For pres~mre vessels, the hydrostatic test pressure "is at least equal to Sy = yield stress, psi
1.3 times the maximum allowable working pressure to be marked on M vL-LL = moment due to dead loads and live loads, in·lb
the vessel multiplied by the lowest ratio (for the materiais of which the Z = section modulus, in3
vessel is constructed) of the stress value S for the test temperatu re to
the stress value S for the design temperature."23 In other words, The pneumatic test pressure is
4.14 Test Pressure for Process Piping

where Pwst = test pressure, psi
Pde"'lf" = design pressure, psi
The hydrostatic test pressure of process piping is2 5
St~st - a llowable stress at t est temperat ure, psi
Sr~,·•ign = allowable stress at design temperature, psi n p stest
rhydn> =- 1.5 X design X -S--
design
ASME V111 Div.l does not specify an upper limit oftest pressure, but
the inspector would reject a vessel visibly deformed after testing. For with
negative pr essure or vacuum service, the "test shall be made at a pres-
sure not less than 1.3 times th e difference between normal atmo- . - D
P.wdru X
-s;Sy
spheric pressure and t he minimum design interna i a bsolute 2xt
pressure.":.!3 P.hydro x D + Q.75i M DL +LL s; S.
For pressure vessels, the pneumatic test pressure is 1
4xt Z
136 Chapter Four Fabrication 137
where D pipe diameter, in

t wall thickness, in
S}· = yield stress, psi
MoL+U. = moment due to deadweight and live loads, in·lb
Z - section modulus, in3
If the piping is tested with a vessel,
77% Phydro 5 Phydro with ve55el 5 P,·essel
where Phydro "ith vessel = hydrostatic test pressure of pipe with vessel, Figure 4.22 Hot crack in double submerged are weld.
psi
Phydro - B31.3 hydrostatic test pressure, psi
Pvessel = vessel test pressure, psi.
The pneumatic test pressure is
4.15 Test Pressure for Liquid Pipelines

Hydrocarbon liquid pipelines operating at a hoop stress larger than 20
percent of SMYS (minimum specified yield stress) are tested ata pres-
sure26
Phydro = 1.25 X Pdesign

Figure 4.23 Crack in offset bead.
Because for liquid pipelines the design pressure Pdesign corresponds to
72 percent ofyield, 1.25 Pde"gn corresponds to 1.25 X 0.72 S1· 0.9 Sy.
Pneumatic testing ts only permitted for low-pressure pipelines, those 4.17 Mill and Handling Flaws
operating at a hoop stress of 20 percent or less of SMYS. Mill and shop fabrication of pipe and components is governed by mate-
rial specifications, codes, and owner procurement specifications. The
4.16 Test Pressure for Gas Pipelines controlled conditions in a mill or fabrication shop generally result in
good quality welds, but exceptions do occur, as illustrated in Figs. 4.22
The maximum allowable operating pressure (MAOP) of an installed to 4.25.1 Damage may also occur during handling and shipment, and
pipeline can be established by hydrostatic testing21 therefore the owner should include a receipt inspection, at least ran-
dom, once materiais or components arrive at the field.
MAOP. = P.hydro MAOP =
4
phydro
I 1.25 2.5
4.18 Field Weld Flaws
where MAOP 1 - maximum allowable working pressure in a class 1, A field weld is first visually inspected for workmanship and general
Division 1 location (noncritical) and MAOP4 = maximum allowable condition, and fillet welds rnay be measured for minimum leg length.
working pressure in a class 4 location (most critica!). When performing visual examinations, there are differences of opinions
as to whether heat tint common in a nickel alloy weld affects corrosion

resistance (Fig. 4.26). Recent reports indicate that the heat tinted mate-
rial is more susceptible to corrosion, and the heat tint should be removed
by pickling.27
After visual examinations, welds are inspected using nondestructive
examination (NDE) techniques, following the requirements of the con-
struction code. Examination techniques are addressed in Chap. 6.
Upon completion of examinations, NDE technicians report the pres-
ence or absence of "discontinuities" or "indications." If, upon evalua-
tion, indications exceed the construction code limit, they are labeled
"defects" or "flaws."
Weld defects, illustrated in Figs. 4.27 to 4.36, may be grouped into
three categories.
1. Cracklike discontinuities
Figure 4.24 Hook crack in electric resistance seam weld. • Cracks. Any forro of cracking, at the weld root, body, or crown
(Fig. 4.27).
• Lack offusion. Successive weld passes do not fuse between them-
selves or with the adjacent base metal (Fig. 4.28).
Figure 4.25 Fatigue crack developed dwing shipment.
Figure 4..27 Weld shrink crack.
Figure 4.26 Heat tint in nickel

alloy weld.
Figure 4.28 Radiography of lack of fusion.
140 Chapter Four Fabrlcation 141
• Incomplete penetration. The root pass does not reach the root of
the weld (Figs. 4.29 to 4.31).
• Ouerlap.
• Weld decay. Cracking in sensitized stainless steel welds.
2. Volumetric discontinuities
• Porosity. Air or gas cavities in the weld; appear as a dark spot in
radiography (Fig. 4.32).
• Slag inclusion. Nonmetallic elements entrapped in weld, such as
an oxide, tungsten from an electrode, flux, and the like, may
Agure 4.31 Radiography of incomplete penetralion.
Figure 4.29 I ncomplete penetration at weld root.

A gure 4.32 Radiography of a cluster of poro:sities.
appear as a dark or light spot in radiography, depending on their

density relative to iron (Fig. 4.33).
3. Geometric discontinuities
• Undercut. A groove adjacent to the toe of a weld.
• Incorrect profile. Excessive protrusion on the pipe outer diameter
and excessive projection at the pipe inner diameter (Figs. 4.33
and 4.34).
• Misalignment . The two ends to be welded are not aligned within
the tolerance o f the construction code.
• Excessive reinforcement. Too large cover pass on weld, protruding
<Figs. 4.33 and 4.35).
• B urn-through. See Fig. 4.36.
Are strikes consti tute a concern when welding high-strength steels
or low-alloy (chrome-molybdenum) steels. In these cases, are strikes
Figure 4.30 lncomplete penetration. should be ground out, and then inspected by PT or MT to verify that
Figure 4.33 Slag inclusion and excessive OD and ID

protrusion.
Figure 4.36 Weld burn-through. 1
there are no flaws left behind; the metal may possibly be etched to
ensure removal of heat-affected material; the remaining wall thick-
ness should then be measured and rebuilt if necessary by con trolled
deposition welding.
4.19 Weld Size

The size of pressure boundary welds and attachment welds to the
pressure boundary are specified by design and construction codes,
Figure 4.34 Weld protrusion at and for repairs they are specified by postconstruction and repair
pipe inner diameter. codes. Butt welds are typically full thickness, with some allowance
for mino1· workmanship flaws.
Fillet welds are specified as a functíon of the wall thickness of the
connected parts. For example, the leg (side) of a fillet weld on a
pipelíne branch saddle is required to be 70 to 100 percent of the
header pipe thickness. But, if we carefully look at Fig. 4.37, not hav-
ing X-ray vision, we could have concluded Oooking from outside)
that the fillet weld is quite large, when in reality it is barely holding
the saddle. In this case, the gap is a criticai parameter and should
be controlled.
4.20 Residual Stress

Residual stresses occur when the hot metal is cooled. It can happen in
the mill as the ingot cools down (Fig. 4.38), or in welds as the weld pool
Figure 4.35 Excessive weld reinforcement at pipe outer
diameter. cools down.
Figure 4.39 Mechanical stress

relief.
Figure 4.37 Undersized weld at hot tap.•

ing when a tank, vessel, or pipe is subject to a large tensile test stress
causing a reduction in residual stresses: a mechanical stress relief.
In Chap. 9, fitness-for-service of cracklike flaws, residual stresses
play a crucial role in predicting the behavior of cracks and the poten-
tial for fracture. In fracture analysis it is therefore essential to measure
or estimate accurately the residual stresses at weldjoints, but bow?
4.21 Measuring Residual Stresses

Hole drilling. Strain gage rosettes are placed in a circle of diameter
D, around a spot. A hole 0.4D deep is drilled in that spot, relaxing
residual stresses at that point CFig. 4.40).29 34 The change in stress-
Figure 4.38 Residual stres,::es in aluminum ingot.2B es is recorded to obtain the residual stress at that location.
Variatwns of hole drilling. Stresses can be measured by strain
Residual stresses in welds are caused in severa! ways. gauges or by interferometry (Fig. 4.41).
Residual stresses measurement by X-ray diffraction (Fig. 4.42). Stresses
• Contraction of weld upon cooling
cause strains that change interatomic bond distance d, these changes
• Structural constraints against free contraction can be measured by neutron ar X-ray diffraction, based on Bragg's law,
• Nonlinear temperature profile
À= 2xdxsin9
• Phase change causing volumetric strains
• Weldmg dissimilar materiais where À. = wave length
d = interat omic bond distance
Figure 4.39 illustrates residual stresses in a weld between two e = angle of wave diffraction cone
plates. In this case, in the as-welded condition CAW) the residual stress
is tensile; near yield at the weld centerline, it decreases and becomes Residual stress measurement based on Barkhausen noise. When a
compressive away from the centerline to finally reach zero. If a large ferromagnetic material is magnetized it emits an electrical signal,
tensile stress is applied to the plate, the stress distribution takes the the Barkhausen effect. The signal amplitude increases with the
form of the applied stress line AS. After release of the applied tensile stress in the metal. For an unloaded part this stress is the residual
stress, the residual stress has been significantly reduced to the resid- stress, and the change in stress can be measured by eddy current
ual stress line R. We see here one ofthe advantages ofhydrostatic test- probesJn
Figure 4.42 Residual stresses by

X-ray difTraction.37
4.22 Calculating Residual Stresses

The analytical prediction of weld residual stresses is one of the major
Figure 4.40 Strcsscraft RS-3D hole drilling equipment. 3;
engineering achievements of the 1990s, and the improvements in this
discipline are continuing. In work sponsored by the Pressure Vessel
Research Council, the Material Properties Council Joint Industry
Project, Jll and European industry,39 analytical tools and protocols have
been developed to predict, with accuracy, the residual stress distribu-
tion in a weld joint.
The analysis of residual stresses involves modeling complex ther-
mal-metallurgical stress-coupled processes. Extensive benchmarks
against residual stress measurements and independent work by vari-
ous researchers have shown the predictions to be accurate. The analy-
sis ofresidual stresses starts by entering the heat input, obtained from
the welding procedure, into a three-dimensional heat transfer model to
calculate the temperature profile as a function of time. In the weld
pool, the temperature will of course peak above the melting tempera-
ture, and then cool down in a matter of seconds, depending on the
thickness of the two parts being welded. By then, a second pass will
restart the cycle again.
Within the weld pool, the analysis models the solid-state phase
change (austenite, martensite, ferrite), using thermal-metallurgi-
Figure 4.41 Hole drilling with interferometry.as
cal-thermoplastic relationships. And this is repeated for each pass
(Fig. 4.43). Correctly simulati ng the multipass process (the remelt-
Fabrication 149
148 Chapter Four
ing process) is one of the more complex features of residual stress

analysis. The phase transformations and the corresponding changes
in hardness (e.g., between martensite and ferrite) can also be mod-
eled and correlate well with hardness measurements. In practice
this means that temper bead repairs can be modeled and weld param-
eters can be selected and qualified to help prepare the repair weld
procedure.
• Figure 4.44 shows residual stresses in a stainless steel weld, wall
thickness 0.65 in and Dlt - 50, three weld passes; tension on bot-
tom and compression on topo f weld centerline, balanced by reverse
stress distribution on both sides of the weld.
• Figure 4.45 shows residual stresses in a stainless steel weld, wall
thickness 1.6 in and Dlt- 200, nineteen weld passes; tension at top Figure 4.45 Stainless steel Dlt · 200.-l'
and bottom, compression in the middle, stresses self-equilibrate

through the weld section.
With the technology of residual stress predictions conquered, there
remained the task of making the information available to the user.
This was achieved in Appendix E of API 579, which provides a com-
pendium of residual stress solutions for common welds.
4.23 Mechanical Joint Flaws

Leaks or ruptures at mechanical joints occur at inadequately bolted
flange joints; inadequately screwed threaded or swayed fittings; at
Figure 4.43 Model of multipa~s inadequate gaskets, seals, and packing; and so on. These shortcomings
weld.40 are breakdowns in engineering. construction, or maintenance. These
are programmatic breakdowns that are not the focus of this book.
These problems can be prevented in severa! ways.
• Goodjoining procedures, in compliance with industry standards and
•• 1111111111111,,,
1IIIIIIIIIII,JIIIII·'
manufacturer requirements and limitations
• Training of construction and maintenance personnel
,,.lli.,
• Preservice pressul'e or leak testing
.....l •••
~········~····-·~
~·····:!•!• i~~~·-~·
~0111' . ~ 1.... - ' ·~
References
1. Kiefner and Associates. Worthington. OH.
2. Courdeuse, L.. et ai. Carborz Manganese Steels for Sou r Seruice- lmprouement of
HIC and SSC Resistance, Pressure Vessel Research Council, Managing lntegrity of
Equipment in Wet H:zS Se rvice, October 10-12. 2001, Houston, TX.
Figure 4.44 Stainless steel Dlt - 50.3' 3. Thielsch Enginet!•;ng. Cranston. RI.
150 Chapter Four Fabrication 151
4. Mohler, R., Practical Welding Technology, Industrial Press, New York. 38. WRC Bulletin 476, Recommendations for deterrnining residual stresses in fitness-
5. AWS ARE-4, Welding Metallurgy, American Welding Society, Miami. for-service assessment, P. Dong, J. K. Hong, November 2002, Welding Research
6. AWS ARE-5, Design for Welding, American Welding Society, Miarrú. Council.
7. AWS ARE-6, Test Methods for Eualuating Welded Joints, American Welding Society, 39. Structural Integrity Assessment Procedures for European lndustry, SINTAP,
Miami. November, 1999.
8. AWS ARE-7, Residual Stress and Distortion, American Welding Society, Miami. 40. Anderson, T. L., lncorporation of Residual Stresses into the API 579 FAD Method,
9. AWS ARE-8, Symbols for Joining and Inspection, American Welding Society, Miami. Welding Research Council, Progress Report. LIX (112), January/February 2004.
10. AWS ARE-lO, Monitoring and Contra/ of Welding and Joining Processes, American
Welding Society, Miami.
11. AWS ARE-11. Mechanized, Automated, and Robotic Welding, American Welding
Society, Miami.
12. AWS ARE-12, Economics ofWelding and Cutting, American Welding Society, Miami.
13. API Std 1104, Weldíng of Pipelines and Related Facilíties, American Petroleum
14. ASME Boiler and Pressure Vessel Code, Section IX, Welding and Brazing
Qualification, American Society of Mechanical Engineers, New York.
15. Sperko, W. •]., Sperko Engineering Services Inc., Greensboro, NC.
16. API Spec. SL, Specification for Line Pipe, American Petroleum Institute, Washington,
DC.
17. APJ RP 2200, Repairing Crude Oil, Liquefied Petro/eum Gas, and Product Pipelines,
American Petroleum Institute, Washington, DC.
18. API 2201, Safe Hot Tapping Practices in the Petmlf'um & PPtrnrhPmirnllnrfustries,
19. Antaki, G. A., Piping and Pipeline Engineering, Dekker, New York.
20. API Standard 653, Tank Inspectíon, Repair, Alteration, and Reconstruction,
American Petroleum lnstitute, Washington, DC.
21. ASME B31.8, Gas TI·ansmission and Distributíon Piping Systems, American Society
of Mechanical Engineers, New York.
22. API Standard 650, Welded Steel Tanks for Oil Storage, American Petroleum
23. ASME Boiler and Pressure Vessel, Section VIII, Div.1 Pressure Vessels, American
Society ofMechanical Engineers, New York, NY.
24. ASME B31.1, Power Piping, American Society ofMechanical Engineers, New York.
25. ASME B31.3, Process Piping, American Society of Mechanical Engineers, New York.
26. ASME B31.4, Pipeline Transportation Systems for Liquid Hydrocarbon,ç and Other
Liquids, American Society of Mechanical Engineers, New York.
27. Nickel Magazine, ,July 2004, 19 (3), The Nickel Institute, 'lbronto.
28. Newbom, M. A., Residual Stress and Measurement in the Alumin.um In.dustry, RS
Summit, December 2003.
29. ASTM E 837, Photoelastic Determination of Residual Stress in a Transparent Glass
Matrix, ASTJVI International, West Conshohocken, PA.
30. ASTM E 837 Determining Residual Stresses by the Hole-drilling Strain-gage
Method, ASTM International, West Conshohocken, PA.
31. ASTM E 915 Verifying the Alignment of X·Ray Diffraction lnstrumentation for
Residual Stress Measurements, ASTM International, West Conshohocken, PA.
32. ASTM E 1426 Determining the Effectiue Elastic Parameter for X-Ray Diffraction
Measurements of Residual Stress, ASTM International, West Conshohocken, PA.
33. ASTM E 1928 Estimating the Residual Circumferential Stress in Straight Thin-
walled Tubing, ASTM International, West Conshohocken, PA.
34. ASTlVI E 2245 Residucd Strain Measurements ofThin, Reflecting Films, SAE, J784a
Residual Stress Measurement by X-Ray Diffraction, ASTM International, West
Conshohocken,PA.
35. Lord, J., Oueruiew of Residual Stress Measurements - A UK Perspectiue, NPL
Materiais Center, December RS Summit, Los Alamos, NM, December 2003.
36. Johnson, E. M., et. al., Residual Stresses in Surface Heat Treated Paris, RS Summit,
Los Alamos, NM, December 2003.
37. Stresstech Group, Stress Measurement by Barkhausen Noise, Stresstech Group
GmbH, Germany, RS Sumrrút, Los Alamos, NM, December 2003.
Chapter
5
Degradation
5.1 Corrosion
Corrosion (ft·om the Latin corodere or "to eat away") is the degradation of
a component by its environment. In a classic textbook on the subject,
C.P. Dillon defines corrosion as ''the deterioration of a material of con-
struction or of its properties as the result of exposure to an environ-
ment."l.2 In its standard terminology related to corrosion, ASTM G 153
defines corrosion as the chemical or electrochemical reaction between
a material, usually a metal, and its environment that produces a dete-
rioration of the material and its properties. Corrosion can be uniform
i f it has proceeded at a uniform rate over a wide area, or local i f it is
confined to small areas.
5.2 The Corrosion Engineer's Perspective

The three key practical questions in corrosion are
• How can corrosion be recognized and categorized?
• Given a metal and an environment, what is the corrosion rate, and
therefore how to predict remaining life, and conversely, how to select
the best metal for a certain service?
• How to best protect against corrosion?
From a corrosion engineer's perspective, corrosion can be categorized
in broad classes. One of the first classifications of corrosion mechanisms
was provided by M.G. Fontana; it consisted of eight categories:
• General (uniform) corrosion
• Localized corrosion
153
154 Chapter Flve Degradation 155
• Galvanic corrosion r-- --

Atmospheric Corrosion
• Cracking phenomena Galvanic Corrosion
Stray-Current Corrosion
• Velocity effects General
General Biological Corrosion
Corrosion
• Intergranular attack Molten Salt Corrosion
Corrosion in Liquid Metais
• Dealloying High-Temperature Corrosion
• High-temperature corrosion
These eight categories were further grouped by C.P. Dillon1·2 and Filiform Corrosion
illustrated in Fig. 5.1; in this case, corrosion mechanisms are classified Crevice Corrosion
Local
in three groups based on the detection method. One more example of Corrosion Pitting Corrosion
corrosion classification is presented in the ASM Handbook4 ·5 and is Localized Biological Corrosion
illustrated in Fig. 5.2.
5.3 The Facility Engineer's Perspective

The facility engineer is oftentimes more interested in the bottom lme: lntergranular Corrosion
Metallurgically
I detected corrosion, how much longer can I operate the equipment lnfluenced Grooving Corrosion
without risk ofleakage or failure? In other words, is the equipment fit- Corrosion Dealloying Corrosion
for-service, and for how much longer?
Erosion
Group I Uniform Corrosion Mechanically Fretting Corrosion
Visual Localized Corrosion Assisted Cavitation Erosion
Examination Galvanic Corrosion Corrosion Water Drop lmpingement
Corrosion Fatigue
Velocity Effects Stress-Corrosion Cracking

Group 11 Environmentally Hydrogen Damage
lntergranular Corrosion
NDE lnduced Cracking Liquid Metal Embrittlement
Dealloying Corrosion
Solid Metallnduced Embrittlement
Figure 5.2 Forms of corrosion based on ASM.
Cracking Phenomena
Group 111
High-Temperature Corrosion For the purpose of fitness-for-service, judging mechanical integrity,
Microscopy
Microbial Effects and making run-or-repair decisions (fitness-for-service), corrosion
mechanisms can be viewed as falling into one o f three large categories:
wall thinning mechanisms, cracking mechanisms, and mechanical
Figure 5.1 Forms of corrosion based on Dillon. degradation.
156 Chapter Five Degradation 157
• Wall thinning can be general, local, galvanic, pitting, crevice, erosion Wall Thinning
(with or without corrosion), or microbial. General
Fi1ness-for-Service
• Cracking can be intergranular or transgranular, cyclic fatigue (with Local
Chapter 7
or without corrosion), hydrogen-induced, or stress corrosion cracking. Pitting
MIC
• Mechanical degradation is degradation that adversely affects mech an-
ical properties, for example, hydrogen embrittlement of steel that
degrades toughness.
Geometric Defect 1----+l Fitness-for-Service
5.4 Damage Chapter 8
In fitness-for-service we are not only interested in degradation by cor-

rosion, we are also interested in the broader question of damage to the
component. This adds two more general categories: high-temperature
effects and geometric defects or damage. Cracking
----.1 Fitness-for-Service
Corrosion lnduced
• High-temperature effects include creep , graphitization, decarburiza- Chapter 9
Fatigue lnduced
tion, high-temperature oxidation, nitridation, temper embrittlement,
and liquid metal embrittlement. The se are classified as damage
hecause creep has traditionally been referred to as "creep damage."
• Geometric defects may be caused during mill and shop fabrication or
High Temperatura Fitness-for-Service
field construction and erection; they include, for example, ovality of a Creep Chapter 10
tank cross section, peaking of a vessel's longitudinal seam, or dent in
a pipeline as it is lowered over a rock at the bottom of the trench.
-
• Geometric damage can occur in service, for example, a dent and gouge Figure 5.3 Degradation and fitness-for-service.
in a pipeline caused by an excavating backhoe, or the accidental bend-
ing of a tank wall or pipe, buckling or rupture of a pipeline due to
soil settlement, or inward buckling of a tank shell dueto accidental • The third and last step is to realize the multitude of parameters that
differential externa] pressure. can influence the onset of corrosion and the corrosion rate.
5.7 The Electrochemical Cell

5.5 Degradation and Fitness-for-Service If we place a steel plate and a copper plate in water, and if we connect
To each degradation mechanism corresponds a fitness-for-service eval- the two plates by a metallic wire, the steel plate will corrode (Fig. 5.4).
uation technique, and a chapter in this book, as described in Fig. 5.3. From the surface of the steel plate, posüive iron ions will dissolve in
the water as ferrous ions
5.6 Understanding Waii-Thlnning
Mechanisms
The iron in steel goes into solution, leaving behind pits and thinned
There are three basic steps in understanding wall-thinning mechanisms. walls. The corroding end (in this case the steel plate) is referred to as
• The first step is to understand electrochemical corrosion in its sim- the anode. At the same time, two electrons will migrate towards the
plest forro: the electrochemical cell. copper plate, through the wire. At the copper-water interface, the
cathode, hydrogen will be formed by the excess electrons
• The second step is to recognize the electrochemical cells in real plant
envi ronments. 2 H ' + 2e ~ 2 H ~ H2
The fonnation of hydrogen blankets the cathode, and polarizes it. In 5.8 The Single Metal Electrochemical Cell
the presence of dissolved oxygen, the cathodic reaction is accelerated
by depolarization In practice, for corrosion to take place, we do not need neatly separated
anodes, cathodes, conductive wires, and electrolytes. Instead, the same
piece of metal will exhibit a multitude of anodes and cathodes (Fig.
5.5). In this case, an externai wire conductor is not needed because the
In aerated water and seawater. the cathodic reaction is electrons can move from anode to cathode within the metal itself. The
last constituent needed is the electrolyte; it could be the fluid inside
Oz 2 HzO + 4e- ~ 4 OH-
the component (internai corrosion), soil, surface condensation, or the
The hydroxide ion OH- will eventually react with the iron ion Fe" humidity in the atmosphere (externai corrosion).
discharged at the anode to fonn severa! fonns of oxides. From the corrod- A decades-old classic experiment is conducted to illustrate electro-
ing steel surface outwards toward the water, we find, in order, greenish chemical corrosion. Pour into a flask containing an aerated electrolyte a
ferrous hydroxide Fe(OHh formed from the reaction small amount of potassium ferric cyanide, which is a reagent that turns
blue in the presence ofFe.. (it becomes blue at an iron anode). Also, pour
Fe·- + 2 OH ~ Fe (OH)z a small amount of phenolphtaleine, a reagent that turns red in the pres-
ence of OH- (it becomes red at the cathode of an aerated electrolyte).
then, black magnetite Fe30 4 n(H20), and then, in contact with the Place a nail into this solution. The body of the nail becomes sur-
water, is red-brown ferric hydroxide fhematite, Fe<OHh], which is rust. rounded by a red halo, and the head of the nail and its tip become
In water service, the outer layer may also include white carbonates blue (Fig. 5.6). The body is cathodic (noble) compared to the head and
and silicates. tip. Yet ali these parts are made of the same metal The difference is
Referring to Fig. 5.4, the metal that corrodes is the anode, the metal dueto the cold-forming process ofthe head and tip. Imagine now this
that does not corrode is the cathode, the wire is the conductor, and the difference in cold working on a much smaller scale, the scale of the
medium that permits the Fe++ ions to go into solution is the electrolyte. metal grain, to understand the multitude of anodes and cathodes in
Focusing on the anode-liquid interface, where Fe +~ goes in to solution, the same metal.
and because the flow of current is, by convention, the direction of
movement of posi tive charges, we can say that corrosion occurs where
current (Fe++) leaves the metal.
To take place, electrochemical corrosion requires four constituents:
anode, cathode, conductor, and electrolyte. Ifthe environment were per-
fectly dry, there would be no electrolyte and no electrochemical corrosion;
this is one reason why gas is d1ied before being transmitted in pipelines.
Figure 5.5 A single metal is a multitude of anodes and cathodes.
Fe Cu
B
B~
R R R
Figure 5.4 The electrochemical ~8 Figure 5.6 Iron nail, blue (B)
anode !corrodes), red (RJ cathode.
cell. R R R
Fe H
8
160 Chapter Flve Degradatlon 161
6~,--:-e--~__:____c_u___:__~:>R
Zn ==::::> Figure 5.8 Zn-Fe nail corrosion

experiment.
R
most easily) to most noble (most cathodic, least corroded), top to bot-
tom and left to right.
Figure 5.7 Jron oxide (rustl from overhead valve has poor adhesion.
Magnesium Stamless steel (active) Nickel alloys (passive)
Zinc Lead Titanium
Another classic case of galvanic corrosion within the same piece of Aluminum Tin Gold, platinum
metal is graphitic corrosion of cast iron in contact with conductive solu- Low-carbon steel Brass, bronze Graphite
tions. Here, the graphite flakes act as cathodes and the surrounding Low-alloy steel Steel mill ~cale Carbon
iron is the anode, which oxidizes in place. In practice, the anodic and Wrought and cast iron Stainlcss steel (passive)
cathodic region8 may shift over time, resulting in evenly distributed
corrosion, referred to as general corrosion. This is illustrated again by the nail experiment in reagent containing
The fonnation of an oxide layer does, for many metais, slow down the solution (Fig. 5.8). If we plate the tip half of the nail with copper (Cu)
electrochemical reaction. Metais that form protective oxides include Be, which is more noble than iron, the nonplated half ofthe nail, head and
Cu, AI, Si, Cr, Mn, Fe, Co, Ni, Pd, Pb, and Ce. Unfortunately, some ofthese body, will turn blue {anodic), and the Cu-covered tip and body will turn
oxides-in particular the iron (carbon steel) oxide-have poor adhesion, red. The anode is the bare iron; the cathode is the copper. Corrosion
andare easily washed away externally by rain or condensation (Fig. 5.7), between two metais in contact, one more noble than the other, is gal-
and internally by the process flow. On the other hand, chromium oxide vanic corrosion.
is very adhesive, which gives stainless steel its appearance and corrosion Finally, if we plate the tip half of the nail with zinc (Zn) which is less
resistance. The chromium oxide is said to have "passivated" the stainless noble than iron, the bare half of the iron nail, head and body, will turn
steel. However, in practice, passivation is more complex: stainless steel red (cathodic). There will be no blue colar anywhere because no Fe++ is
is resistant to oxidizing acids such as nitric acid at room temperature, discharged at the Zn anode. The Zn has sacri:ficed itself in protecting the
but suffers severe pitting in the presence ofhydrochloric acid. steel. Zn has acted as a "sacrificial anode." This is the principie behind
galvanic protection of steel by zinc coating (galvanized steel).
5.9 The Galvanic Cell
5.1 O Concentration Cell
Ifthe anodes and cathodes are fixed, then corrosion will proceed at these
fixed points causing local corrosion and pitting. This is the case if we We have seen that corrosion rates depend on nonuniformity ofthe metal
connect two metals with different potentials in the galvanic series. itself, in the forms of anodes and cathodes. Corrosion rates are also
Following is a list of some metais from least noble {most anodic, corrode affected by nonuniformity of the fluid in contact with the metal. This
Cu Fe
Heavy Light
corrosion corrosion
Figure 5.10 Plat.e-rivet experiment.
5.11 Size Effect

Figure 5.9 Oxygen concentration ccll.
A small anodic region will corrode rapidly when faced with a large
cathodic region. Thi::> ::>i:.~;e dfecL i::> illustrated by another classic exper-
form of corrosion is referred to as concentration cell corrosion. A classic iment (Fig. 5.10). On the left, two copper plates are connected by iron
experiment consists of forming a cell with two pieces of steel connected rivets and immersed in seawater. At the same time, two iron plates are
by a wire and voltmeter, and immersed into an electrolyte (Fig. 5.9). connected by copper rivets and also immersed in seawater, shown on the
Nitrogen is bubbled next to one piece of steel, and oxygen is bubbled right. In the first case, the cathode is large (copper plates) and the anode
next to the other. The side bubbled with nitrogen is starved in oxygen. small (rivets); the reverse is true in the second case. When the plates
The voltmeter will detect the passage of current; the side starved in oxy- are pulled out ofthe sea severa! months later, the iron rivets on the left
gen is anodic and corroding relative to the side rich in oxygen. The are brushed away as easily as wet chalk; they have completely cor-
"oxygen concentration cell" has caused one steel piece to corrode. roded: the small anode (iron rivets) has corroded ata high rate when
This helps explain what happens inside crevices, for example, in the in contact with the large cathode (copper plate). On the contrary, the
interstice of threaded joints, or at attachments between internais and iron plates on the right show only minor signs of corrosion: the large
vessels, or at tank bottom edges. Oxygen is depleted in these tight anode (iron plates) has corroded only slightly when in contact with the
areas and they become anodic compareci to the surrounding environ- small cathode (the copper rivets). This size effect helps explain what
ment. In steel, the iron íons dissolve as we have seen earlier happens in pitting corrosion where small anodic pits are surrounded
by a large cathodic metal.
Fe--? Fe++ + 2e-
More generally, with any metal M, the reaction is 5.12 Parameters Affecting Corrosion Rate
Understanding what caused corrosion and predicting the future corro-
M--? M+ +e-
sion rate is essential to fitness-for-service and run-or-repair decisions.
But, in the absence of oxygen, the metal oxide does not form; instead, It is also, arguably, the most difficult aspect offitness-for-service. Even
negative íons in the solution migrate towards the M+ rich crevice to form in closely controlled laboratory tests, and in the simple case of uniform
an acidic environment. For example, in the presence of chlorides, hydro- corrosion, corrosion rates of duplicate specimens are at best within ::t:: 10
chloric acid is formed; in the presence of sulfates, sulfuric acid is formed, percent of each other.7 The primary reason that corrosion rates are dif-
with the pH dropping to nearly 1 or 2 in the crevice6 ficult to predict is that corrosion depends on such a large number ofvari-
ables, and these variables vary during service and shutdowns. These
M+Cr + H20 ~ MOH J- + HCl variables include
Mz+SO::j- + 2H20 ~ 2MOHJ- + H2S04 • The fluid. The electrolyte, the ions in solution.
• The atmosphere. The temperature, humidity, or corrosiveness of the • The presence ofa weld. Because the weld or heat-affected zone can be
atmosphere or soil. anodic or cathodic.
• The temperature. Corrosion rates are temperature dependent because • The fabrication method. Castings generally corrode more readily than
ionization, reaction rates, and diffusion rates increase with tempera- equivalent wrought metal.
ture. In some cases an increase of20"F could double the corrosion rate.
There are exceptions, such as zinc that passivates in high-temperature
water. Also, boiling will uniformly reduce the oxygen content of the 5.13 Predicting Corrosion Rate-ls lt Linear?
liquid, which is why steel corrodes less in boiling water.ll Another Corrosion rates can be measured in the laboratory or in the field, as
example is a nickel-base alloy in aerated sulfuric acid: the corrosion described in Chap. 12. But Fig. 5.11 illustrates the fact that the corro-
rate varies frorn 30 rnpy at 80°F, up to 120 mpy at 180°F, and back sion rate, are not necessarily linear. In this case, three steel plates
down to practically zero at 210cF.4.5 were placed on racks on a beach and exposed to a marine environment
• The concentration. Corrosion rates are dependent on íon concentration for severa! years. One was ordinary steel, another had low copper, and
in the fluid or environment. The relationship is not linear. For exaro- the third one was an Ni-Cu steel. The weight of each plate was period-
pie, the corrosion rate of a nickel-base alloy is 90 mpy in 10% sulfuric ically measured to establish corrosion rates over a period of severa}
acid, bui then drops to 40 mpy in 60% sulfuric acid, to climb back up years. The data are presented in Fig. 5.11. The corrosion rate ofthe low
to 140 mpy in soe~ sulfuric acid4.5 copper and t.hP r.n-Ni Rt.PPlR waR clP.arly nonlinear the first two years.
• The flow velocity. Corrosion may be accompanied by erosion. In the case ofFig. 5.11, the environment was the same for all three
metal plate specimens: seaside, with the cycle of days and nights, and
• The amount of dtssolued oxygen. The complete elimination of oxygen seasons. In a plant, things are more complicated: the environment varies
will reduce the corrosion rate, but ifthe tluid contains some oxygen, the with the flow stream (on purpose or by accident) and the extemal con-
areal:> starved in oxygen become anodic, as illustrated in Sec. 5.10. ditions vary (e.g., corrosion under insulation, acids in the atmosphere
• The phase ofthe fiuid. Liquid, vapor, or gas. from a leak, etc.). As a result, the actual corrosion rate is even less linear
• The pH of lhe solution. Graphs plotting metal-solution potential ver- than shown m Fig. 5.11.
sus the solution pH, ata certain temperature, are called Pourbaix Despite all these difficulties, the practicing engineer needs to estimate
diagrams. corrosion rates, on the basis o f a nominal environment, and one of the
best references for generic corrosion rates is NACE's Corrosion Survey
• The contaminants in the flow stream. Afew parts per million (ppm) of
Database.9
ferric ions F e+++ in solution cause the accelerated corrosion of stainless
steel coils. The ferric ions act as oxidizing agents, taking electrons
away from the steel surface Fe+++ +e---+ Fe++.
• The procest; conditwns. Operation, shutdown, wash, and so on. 80
•• Metallurgy (hardness, cold work, grain size). A cold-worked metal is 70 •
more anodic than the same metal annealed. Grain boundary atoms #. 60
are notas strongly bonded and tend to corrode more readily. vi
~ 50
• The weld properties. Heat treatment, hardness, residual stresses, sen-
sitization IIAZ, inclusions. .E
Ol
40
• The component geomet1y. Crevkes, local turbulence. ~ 30
• The coating and lining condition. Holidays, disbondment. 20
• The reiative size of anodic and cathodic regions. 10

Figure 5.11 Corrosion rate of
• The solubility of corroszon products. steels in manne environment.
o 1o 20 30 40 50 60 70 80 90
• The addition of corrosion inhibitors. Type, quantity, and distribution. Months
5.14 Predicting Corrosion Rate- Some of these deposits may have originated upstream and been carried
Time in Service down by the flow. These deposits and the underlying corrosion pits are
also referred to as tubercules.
Corrosion rate is simply total corrosion divided by time in service, but
There are two primary problems related to the formation oftubercules.
beware of errors in estimating time in service. In one example, flow-
accelerated corrosion (FAC) caused the catastrophic failure of a turbine • Corrosion rates tend to increase inside tubercules due to the formation
steam extraction pipe in a power plant. The ASTM A 106 Grade B car- of acidic conditions and possibly microbial attack. Permeable deposits
bon steel pipe operated at 300 psi and 425°F. The failure triggered a can lead to "concentration of a corrodent in the stagnant solution,
critica} review ofthe plant inspection program. Earlier analyses of cor- which can be 10 to 100 times or more larger than that measured in a
rosion locations indicated that bends and elbows had a higher likelihood bulk fluid."6
offailure, so these components were inspected and corrosion rates were
calculated on the basis of the plant operating life: 14 years. In reality, • Tubercules reduce flow area and impair heat transfer. In many cases
the plant operating life, 14 years, was not the system's operating life: this is how they are discovered, as the flow rate decreases in the
the parts inspected had been replaced 2 years earlier. The real wear rate system.
was therefore 14/2 = 7 times larger than assumed in the corrosion rate
prediction.1o
5.16 General Corrosion
General corrosion (also referred to as uniform corrosion) is a corrosion
5.15 Deposlts and Tubercu lation
process that results in uniform wall thinning. It is typical when the
When inspecting the inside surface of corroded components, keep in mind same metal contains evenly distributed anodes and cathodes. Uniform
that we find not only corrosion products, for example, ferric oxide (Fe 20 3 ), coiTosion is common on the surface of bare carbon steel in humid
hydrous ferroxide (Fe(0H)3), and magnetite (Fe 3 0 4 ) on steel, but also con- atmospheres, where the corrosion product (rust) is washed away by
taminants, fluid-borne minerais, and treatment chemicals (Fig. 5.12). rain or condensation (Fig. 5.13).
General corrosion rates can be measured by exposing metal speci-
mens to the environment. For example, to study corrosion in a marine
Figure 5.12 Accumulation of deposits. Figure 5.13 General corrosion of bli nd flange.
environment, metal plates are mounted on racks by the seaside and

left exposed to the elements for a period of time. 11 After exposure, the
plates are cleaned, dried, and weighed. The weight loss is then con-
verted to a corrosion rate 12
CR=3.45x106 A W
xTxD
where CR = corrosion rate, mils/year (mpy)
W = mass loss, g
A = area, cm2
T = time of exposure, h
D = density, g/cm3
Corrosion rates can also be estimated based on the measured corro-

sion current between the anode and cathode. For steel, Fe corroding
into Fe++
CR = 45.7 X i
where CR = corrosion rate, mils/year (mpy), and i = corrosion current

density, A/m2 •
Figure 5.1 4 Acid leakage out-of-vessel head llange.
For example, if a laboratory cell that reflects the anode-cathode and
electrolyte of the field condition measures a corrosion current density
of 0.5 A/m2, then the corrosion rate will be
the vessel head flange joint and corroded the carbon steel. In this case,
CR = 45.7 x 0.5 = 23 mpy it was estimated that one drop per second (approximately 0.001
gal/min) left. a deposit buildup of nearly 15 lb of acid in one year.
In practice, a corrosion rate below 3 mpy can typically be accommodated
by a reasonable corrosion allowance of X6 in for 20 yr, or ~ in for 40 yr.
At the other extreme, a corrosion rate over 30 mpy is extremely large 5.17 Galvanic Corrosion
and should be avoided by selecting different alloys or modifying the Galvanic corrosion occurs at a contact interface between a noble meta]
process. Corrosion rates are not usually constant over time. and one less noble on the galvanic series, for example, copper (noble) and
Corrosion products can slow down the corrosion process and, ifthey are carbon steel Oess noble). The galvanic contact is in most cases uninten-
passivating and adhere to the surface, they could stop the corrosion tional, such as in the case when copper from heat exchanger tubes
process altogether. erodes away and settles in downstream carbon steel pipe.
In order to reduce costs, vessels in corrosive service are often fabri-
cated out of carbon steel and then clad with an alloy liner on the inside.
5.18 Pitting
This approach has proven useful and of course less costly than fabri-
cating the whole vessel out of a corrosion-resistant alloy. A problem Pitting is a local, nearly hemispherical, or deep and narrow loss of
arises if the corrosive fluid leaks out and corrodes the vessel's carbon metal, at times widespread and overlapping, often resulting in pinhole
steel outer surface. An example of such externai corrosion is shown in leaks. The environment tends to be acidic inside the pit and therefore
Fig. 5.14. This is a carbon steel pressure vessel lined with st ainless could have a higher corrosion rate than the rest of the component. Most
steel in acid service. After several years of service, acid leaked out of pitting is generalized, affecting a large zone of the component's inner
Degradatlon 171
170 Chapter Flve
Figure 5.15 Pitted outside wall of bare buried pipe.
or outer surface. Figures 5.15 and 5.16 are examples ofwidespread pit-
ting on the outside diameter of a bare buried pipe. In Fig. 5.15, pitting
has caused a pinhole leak that is repaired by a welded patch.
There are severa! methods to measure and characterize the extent
Figure 5.16 Decp pitting of pipc in casing.
of pitting.Ia Pitting can be inspected visually and assigned a density
using standard charts. Pit size can also be measured individually or
statistically. Large pits, such as those at coating holidays in buried
pipelines, can be measured by micrometers or depth gauges (Fig. 5.17 ).
Lasers are used effectively to map a wh ole region of pits (Fig. 5.18).
Small pits can be measured precisely with a microscope at 50X to 400X
magnification. Finally, metallography may be used to study the exact
relationship between pitting and the local microstructure.
5.19 Crevice Corrosion

Crevice corrosion <Fig. 5.19) can occur when "a crevice limits access of
the bulk environment to a localized area of the metal surface."14 The
crevice can be caused by the shape of the component, deposits, or break-
downs in coating. 15 A related corrosion mechanism is deposit corrosion
which is localized corrosion under or around a deposit or collection of
material on a metal surface. Figure 5.17 Measurement of pit dcpth wíth pit gauge.
Figure 5.18 Pit measurement by laser mapping.
Figure 5.20 Local corrosion under disbonded coating.
5.20 Corrosion under lnsulation

Much damage in process and power plant tanks, vessels, and piping is
caused by corrosion under insulation <Fig. 5.21). Liquid seeps under
the insulation and stays there, stagnant, causing an electrochemical
cell in a humid crevice environment. Furthermore, if the insulation
contains chlorides they will be washed against the metal. The source
of wetness under insulation may be as follows.
• Rainwater
Figure 5.1 9 Corrosion in crevices. • Condensation of atmospheric humidity
• Process fluid
The local stagnant environment has a different chemistry than the
bulk fluid; for example, it may be depleted of oxygen in an otherwise Conditions that contt;bute to corrosion under insulation (CUI) are as
oxygen-rich fluid, or it may be more acidic or contain higher chlorides follows .17-20
than the bulk fluid. The crevice becomes anodic rel ative to the metal
Operating temperature:
exposed to the bulk fluid, causing COITosion to proceed. Crevice corrosion
depends on the fluid and crevice chemistry, the area ratio between the • From 120 to 212°F (boiling)
crevice (anodic) and the exposed part (cathodic), and the bulk flow rate. • Operation (metal temperature) below dew point (accumulation of
Crevice corrosion can be replicated and measured in the laboratory or in condensate from ambient humidity)
the field.ll.l4 • Cool dead-legs not heated by process stream (condensation)
An example of crevice corrosion under disbonded coating is illustrated Insulation:
in Fig. 5.20. The expanding volume of the iron oxide further peels away • Water absorbent
the paint coat. Filiform corrosion is corrosion under coatings in the form • Chemistry incompatible with pipe (e.g., chlorides in insulation of
of random filaments. stainless)
174 Chapter Five Degradatlon 175
• Steam tracing that could leak

• Tropical
• Marine
• Signs of dirt or neglect
• Heavy vegetation around jacket
Inspection history:
• Signs of corrosion under insulation in this or similar systems
5.21 Liquid-Line Corrosion

Surface tension slightly raises the liquid level around the edges of the
liquid-air or liquid-vapor line (Fig. 5.22). This thin meniscus in contact
with air is rich in oxygen and the metal behind the meniscus of liquid
is cathodic relative to the metal right undemeath, which has less oxygen,
and which therefore will corrode more rapidly (Fig. 5.23).
5.22 Microbial-lnduced Corrosion

Microbial corrosion, microbial-induced corrosion, or microbiologically
influenced corrosion (MIC) are synonymous expressions that refer to
corrosion caused or accelerated by living organisms.3 •16
Figure 5.21 Corrosion under storage tank insulation. Several species of microorganisms contribute to corrosion:
• Damaged, open seams or missing insulation • By forming crevices

• Separated caulking • By forming concentration cells, for example, locally depleting oxygen
Coating: • By trapping corrosion products
• Poorly applied or deteriorated coating or painting • By concentrating halides, mineral acids, ammonia, or hydrogen sulfides
• Coating over 15 years old
lnsulation jacket (lagging):
• Openings in jacket at hangers, branch points, valves and fittings,
ladders, platforms, flanges, and so on
• Open jacket seam, oriented upward
• Missing bands on jackets
• Inspection ports or plugs, openings
• Signs of rust
Location:
• Six o'clock (bottom of pipe)
• Low point on vertical runs
• Near nozzles or externai attachments Figure 5.22 Liquid-Jine oxygen
concentration cell.
Ambient:
• Areas exposed to mist (cooling towers, process vapors, etc.)
176 Chapter Five Degradat lon 177
sample, characterize, and count microbial colonies; they include micro-

scopic evaluation, imrnunochemical analysis, culture, and biochemical
methods.23.24 In the presence of oxygen in the liquid (aerobic MIC), iron-
oxidizing bacteria catalyze the reaction of iron with oxygen to form
insoluble ferric oxide Fe20 3 . The fen-ic oxide can accumulate to the point
of restricting flow.
In many electrochemical corrosion reactions, hydrogen evolves at the
cathode, and-over time-blankets the cathodic region, slowing down
the con·osion process. However, some bacteria consume hydrogen formed
at cathodic regions; this elimination ofhydrogen depolarizes the cathode,
maintaining a high corrosion rate.
Sulfate-reducing bacteria catalyze the depletion of oxygen and the
formation of sulfuric acid
The acid will in turn corrode the iron, forming íron sulfides,
Figure 5.23 Liquid-line corrosion. 5.23 MIC Prevention

To prevent MIC one must avoid using untreated water under stagnant
warm conditions. This is achieved by keeping systems drained and dry
when not in service during extended lay-up or shutdown, and by pro-
viding intermittent flow or flushes in dead-legs or stagnant tanks. A
pipe slope of Yíti in to X in per foot, and sufficient vent and drain points
will facilitate draining and drying operations.
Biocides are comrnonly used to prevent MIC. The biocide is selected
based on tests of microorganism cultures, chemical and PH tests of fluid,
Figure 5.24 MIC in buried pipe.2"
analysis of the metal surface, and soil analysis in the case of externai
MIC in buried pipe. Note that the choice ofbiocide must be compatible
with the metal; for example, chlorine may be an effective biocide but will
• By forrning acids attack stamless steel.
• By destruction of coatings Bactena attack rnost metais, \vith few exceptions such as titanium
and copper alloys. But, if steel must be used, austenitic stainless steel
The end effect ofMIC is wall thinning in the form ofpitting (Fig. 5.24) with 6% rnolybdenum has provento be more resistant to MIC.
and accurnulation of corrosion products, causing a reduction in flow Cement or polyester linings, in good conditions, have proven to be
area. MIC is prevalent in cooling water systerns, aqueous waste treat- effective in prcventing MIC in water lines, by shielding the metal surface
ment systems, and systerns containing nearly stagnant water, particu- from the organisms.
larly in stagnant or low-flow systems and at warm temperatures (70 to
120°FJ.21 5.24 MIC Mitigation
There is a multitude ofbacteria species that can corrode metais: some
are anaerobic (do not require oxygen) and others are aerobic (require Removing MIC once it has taken place requires cleaning to remove
oxygenl. Several quantitative methods are used by microbiologists to bacteria colonies attached to the component wall, followed by preven-
tion, such as treatment with biocides. However, there are at least two inner diameter, amidst less corroded peaks (mesas), and the corrosion
situations where the component, and possibly the whole system, must mechanism is referred to as mesa corrosion.
be replaced.
• If MIC attack h as been significant, to the point of restricting flow or 5.26 Erosion
causing severe pitting or leakage Erosion is the loss of material due to wear caused by the moving fluid
• If MIC deposits adhere to the metal, cannot be washed away, and (liquid, vapor) or suspended solids. Related to the erosion mechanism
therefore shield the microbial colonies from the bulk flow containing is erosion-corrosion, a conjoint action involving corrosion and erosion
the biocide in the presence of a moving corrosive fluid, leading to the accelerated
loss of material. Impingement corrosion is a forro of erosion-corrosion
Strongly adhering microorganisms and large organisms, such as generally associated with the local impingement of a high-velocity,
mussels and mollusks that accumulate in intake structures of cooling flowing fluid, or suspended solids, against a solid surface.3 Fretting is
water systems, can be eliminated by periodic cleaning and mechanical another form of erosion, caused by oscillatory friction between the two
removal. This is achieved by scraping or jetting with pressmized water surfaces.
or air, and by chemical treatment with bleach or chlorine. Erosion tends to occur under the following conditions.
5.25 Carbon Dioxide Corrosion • High liquid velocity and turbulence, for example, within 10 diameters
downstream of elbows, tees, orifices, or control valves. Typical erosion-
Steel corrodes in the presence of carbon dioxide (C02) in solution with cotTosion rates of carbon steel in seawater vary from 6 mpy at 1 ftls to
water. In the oil and gas pipeline industry this forro of corrosion is called over 40 mpy at 30 ftls.
sweet corrosion (as opposed to sour corrosion dueto H2S). Carbon diox-
• Impingement areas, for example, impingement from internai discharge
ide in solution in water will form carbonic acid,
nozzles on vessel or tank internais
• Entrained solids in liquid or gas
The carbonic acid dissociates into íons of hydrogen and bicarbonate, • Condensate droplets entrained in vapor
• Bubble collapse dueto cavitation, for example, sudden pressure drop
H2COa -t H• + HCOa-
at an orifice plate or a control valve, causing the liquid to vaporize,
The acidic conditions caused by the formation of H+ promote the and the vapor bubbles to collapse once the pressure is reestablished,
cathodic reaction and the formation of hydrogen, within a couple of diameters of the orifice.
H++ e- -t H. To prevent erosion and corrosion, carbon steel components may be

clad with a layer of material more resistant to erosion, such as Hastelloy
At the same time, at the anode, iron goes into solution and combines C, titanium, Monel, Stellite (Co-Cr), or stainless steel. However, if the
with the bicarbonate ions to form a light brown layer of siderite that stainless steel is sensitized during the cladding process, it becomes itself
protects the wall, reducing the corrosion rate. subject to intergranular stress corrosion cracking (Fig. 5.25).
The corrosion rate depends on several factors.
• The partial pressure of carbon dioxide, with high corrosion rates if 5.27 Cavitation
the carbon dioxide (C02) partial pressure exceeds 30 psia.
Cavitation is the forroation of vapor bubbles as a result of a local pres-
• The presence of water, for example, at the bottom of low-velocity oil sure drop, and the subsequent collapse of the bubbles when the pres-
pipelines (in the order of 2 ft/s or less). sure is reestablished. The continuous collapse of millions of tiny
• The stability of the siderite layer. Where the siderite is lost, for exam- bubbles against the inner wall of the component eventually causes the
ple, by erosion, corrosion will continue, forming pits (valleys) on the wall to erode. Typical examples are pressure drops downstream of orifice
Figure 5.27 Erosion of an elbow by saturated steam.

Figure 5.25 SCC in stainless claddmg.u
5.28 Vapor-Liquid Erosion

Wet steam erosion-corrosion is due to water droplets entrained by the
steam and impinging on the metal. The factors that affect wet steam
erosion-corrosion are as follows.2s
• The flow path. Impingement velocities are larger at elbows (extra-
dos), tees, downstream of orifices, and downstream of flow control
valves.
• The alloy content. Carbon steel is more easily eroded-corroded than
low-alloy steel or stainless steel.
• The moisture content ofuapor or gas flow.
Saturated steam is a classic example of water droplets entrained by
the steam flow. Erosion takes place by impact of the liquid droplets
Figure 5.26 Erosion of pump casing. against the internai surface of pipes, particularly at changes in direction
(Fig. 5.27).
plates or control valves, or cavitation in pump casings with insufficient

suction pressure (Fig. 5.26). As the bubbles collapse near the metal 5.29 Erosion in Gas-Liquid Service
surface, the liquid impinges on the surface at high velocities (in the Recommended practice API RP-14E27 provides a criticai velocity for the
order of 300 to 1600 fUs).6 onset of erosion in gas-liquid service
Cavitation damage produces heavy pitting, with a very clean surface
because deposits are removed by the cavitation impingement process.
Hard materiais are resistant to cavitation; they include Stellite, cobalt,
stainless steel, and nickel-aluminum alloys, or ceramic-containing
linings. where v = velocity at onset of erosion, ft/s
c pipe material constant, 100 to 200 for steel in no-solids,

=
degradation appears as deep and large pits among a zone of general
noncorrosive Cor inhibited) service metal loss.
p = density of liquid-gas mixture, lb/ft
3 Severa! accidents due to erosion-corrosion, also referred to as flow-
accelerated corrosion (FAC), have occurred in power plants.28-33 In one
12409xS1 xP+2.7 xRxS xP case, an 18-in elbow off a 24-in feedwater header (both carbon steel)
p= g ruptured, causing a 2-ft X 3-ft section of the pipe to blow out, and
198.7 X p X R X T X z
thrust forces at the rupture whipped the fractured pipe a distance of
where P operating pressure, psia about 6 ft. In another case, a 24-in carbon steel elbow in a power plant's
S1 liquid specific gravity (water = 1; use average gravity for steam extraction line ruptured, with an opening of approxímately 2 ft X
hydrocarbon-water mixtures) at standard conditions 2 ft. In this case, the 375-mil nominal wall had thinned down to 17 mils.
R - gas-liquid ratio, re/bbl at standard conditions An ultrasonic inspection only four months earlier had failed to detect
T = operating temperature, o R this thinnest location. Erosion-corrosion h as been extensively studied in
Sg - gas specific gravity (air = 1) at standard conditions the power industry by the Electric Power Research Institute (EPRI) and
Z = gas compressibility factor, dimensionless by utilities. In the power industry, erosion-corrosion is divided into single-
phase er osion-corrosion and wet steam erosion-corrosion.32
The parameters that affect single-phase erosion-corrosion in carbon
5.30 Liquid Pipelines and low-alloy steel are as follows.3L32
The velocity above which erosion can take place in liquid pipelines • A flow velocity above 5 ft/s for carbon steel. Flow velocity of 15 ft/s or
is given by27
more should be of concern with all materiais.
• A low pH, below 9 for steel. In these cases, titanium has been a good
substitute to steel or copper alloy.
• An oxygen content below 50 JJ.g/kg or 50 ppb.
where v
3
= velocity at onset of erosion, m/s, and p = density of liquid, • A temperature near 300°F. Below and above this temperature the wear
kg/m. rate appears to be lower. For example, for carbon steel, a wear rate of
1 JJ.g/cm2·h at 100°F, increases to 3000 JJ.g/cm2·h at 300°F.
• The flow path. Th e more turbulence, the more wallloss. For example,
5.31 Liquid-Sand Pipelines common locations of erosion-corrosion in carbon steel water and steam
The criticai velocity at which erosion rate could exceed 10 mpy is27 system s are elbows (extrados, i.e., the outside curve ofthe elbow), tees,
and downstream of orifices.
4D • A low-alloy content. The wear rate is worst in carbon steel, and
v= JW
improves with low-alloy and stainless steels. Chromium, molybdenum,
and copper are useful alloys in reducing erosion-corrosion. For exam-
where v velocity at 10 mpy erosion, ft/s ple, l X Cr-X Mo steels (grade Pll) and 2X Cr-1 Mo (grade P22) have
D internai diameter, in shown marked improvement in wear rate compared to plain carbon
W = sand p1·oduction rate, bbllmonth (1 bbl = 945 lbs) steel.
If the consequence of pipe failure by erosion-corrosion would be
5.32 Erosion-Corrosion
unacceptable, a large number of pipe locations, and at least 2-ft long
Erosion due to flow turbulence, cavitation, or impingement can be zones at each location, should be inspected periodically, because the
accompanied by corrosion. If the flow ís intermittent, the cotTosion exact location where erosion-corrosion can occur is difficult to pre-
buíldup can be significant during the periods of no-flow. This type of dict. Experience indeed indicates that some pipes failed only a few
184 Chapter Fíve Degradatlon 185
feet away from where earlier inspections showed no significant wall

loss.
5.33 Environmentally Assisted

Cracking Mechanisms
Environmentally assisted cracking refers to "the initiation or accelera-
tion of a cracking process due to the conjoint action of a chemical envi-
ronment and tensile stress."3 The tensile stress tends to open the crack,
hence the name stress corrosion cracking (SCC). The stress can be due to
applied loads or it may be a residual stress from fabrication and erection.
Cracking can proceed preferentially along grain boundaries (inter-
granular), or cut through grains (transgranular). Intercrystalline or trans-
crystalline corrosion can also take place without either applied or
residual stress, in which case it may simply be referred to as corrosion
crack.ing.
The onset and progression of 8CC are difficult to predict, but can be
reproduced in laboratory testing. Stress corrosion testing is commonly
conducted on metal strips that are bent and held in a U-shape by a bolt.
SCC, ifit does occur, will initiate at the outer face ofthe bent strip, which Figure 5.29 Optical micrograph intergranular cracking.
is in tension. For most environments the cracks are intergranular;
they follow the grain boundary, as illustrated in the scanning electron
microscope micrograph of Fig. 5.28, and the optical microscope micro-
Figure 5.30 CroRs section of stress corrosíon cracked wall.22
graph of Fig. 5.29. In some cases, such as chloride stress cracking of

unsensitized stainless stee] the cracks may be transgranular (cutting
through the grain).
There are many metal-environment combinations that can lead to
SCC (Figs. 5.30 and 5.31). For example, generally:
• Carbon steel in concentrated caustic, nitrates, anhydrous ammonia,
nitric acid, carbonate solutions, and hydrogen sulfide
• Stainless steel in chlorides, hot caustics, oxygenated water, hydro-
gen sulfides, but not ammonia
• Nickel alloys in hot caustics, high-temperature steam
• Copper alloys in ammonia, amines, sulfur dioxide, nitric acid, but not
chlorides
Figure 5.28 SEM micrograph intergranular cracking. • Titanium alloys in nitric acid, and chlorinated hydrocarbons
Figure 5.31 Stress corrosion cracking through asphalt coating.n
Figure 5.32 Knifeline attack peels

5.34 Corrosion Fatigue off a longitudinal seam weld.
Fatigue is cracking caused by cyclic tensile stresses. The cyclic stresses
can be dueto mechanicalloads (such as vibration in service) or thermal
cycling (such as the continuous mixing of cold and hot streams). Crack-
ing proceeds perpendicularly to the tensile stress and is therefore trans- of chromium which is now hostage to the carbon. The grain boundaries
granular. Corrosion fatigue is fatigue in a corrosive environment and have become zones with little chromium that are surrounded by the
occurs at lower stress leveis and progresses ata faster rate of fatigue chromium-rich stainless steel matrix. The stainless steel is said to have
in a noncorrosive environment. been sensitized. In corrosive environments, such as oxalic acid and nitric
Some metais, such as carbon steel and titanium, have an endurance acid, this situation will cause preferential attack along the chromium-
limit, a stress below which fatigue cracking will not occur. The endur- depleted grain boundaries. This type of attack, when it occurs along
ance limit is in the order of 30 percent of the ultimate strength of the the heat-affected zone of welds, is referred to as a knifeline attack, and
metal. Other metais, such as stainless steels and aluminum, do not is illustrated in Fig. 5.32 where a longitudinal seam weld in stainless
exhibit an endurance limit. steel pipe was so severely attacked along the heat-affected zone that
Environmental-assisted cracking mechanisms are complex phenom- the weld peeled off. Such intergranular corrosion of the sensitized
ena sensitive to small changes in metallurgy, stress (including residual heat-affected zone of welds is also called weld decay.
stresses, which greatly complicate matters), and environment. One good When in the early 1980s some power plants developed intergranular
starting point for corrosion rate is a compendium such as ASM's Atlas stress corrosion cracking in 304 and 316 austenitic stainless steel pipes,
of Stress Corrosion and Corrosion Fatigue Curves,34 which provides a parallel solutions were implemented. In a first approach, the cracks were
large family of curves, grouped by type of metal and alloy, and looks at reinforced with weld overlay or with steel sleeves welded around the
crack growth as a function of a multitude of parameters such as envi- pipes. In a second approach, the 304 and 316 steel was replaced by spe-
ronment, applied stress or stress intensity, concentration, cycles to cial grade steels with very low carbon, sulfur, and phosphorus. The
failure, hardness, temperature, partial pressure, and the like. new alloys worked well but were more difficult to weld.
The following options may apply to reduce the risk of stress corrosion
cracking in stainless steel:
5.35 Sensitized Stainless Steel
• Use a low-carbon alloy, such as 304L or 316L.
In the temperature range of 900 to 1500°F, stainless steel becomes "sen-
sitized," which means that it becomes susceptible to intergranular cor- • Quench the white-hot metal soas to pass very quickly through the
rosion or cracking in specific environments. This happens because, in sensítization zone between 1500 and 900°F.
this temperature range the chromium in the stainless steel diffuses to • Stabilize the steel with titanium (type 321 stainless steel) or colum-
the grain boundary where it combines with carbon atoms and precipi- bium (type 347 stainless steel) which will precipitate preferentially
tate as chromium carbide Cr23C6 • The grain boundaries are depleted with carbon, leaving the chromium in place.
Figure 5.33 Corrosion testing of stainless steel in boiling acid.
• Procure materiallots that have been tested for sensitivity to stress

corrosion cracking (Fig. 5.33). A lot of stainless steel components or
plates can be tested for susceptibility to intergranular attack by Fig ure 5.34 Aftermath of the Lemont Refinery explosion and fire, 1984.
immersion in boiling solutions of these acids. 35 lf the metal shows
signs of intergranular attack, the lot is discarded. Pitting of steel in wet H 2S service is the result of the reaction of sul-
fides with iron to produce iron sulfides (such as pyrite FeS2 or smythite
5.36 Sour Corrosíon Fe3S 4) that settle on the inner surfaces of tanks, vessels, and pipe. If
the iron sulfides are near bare metal they form a galvanic couple that
An example of stress corrosion cracking is sour corrosion. lt is cotTosion further corrodes the bare steel.
that takes place in environments containing hydrogen sulfide H2S. It Sulfide stress cracking (SSC) occurs under the combined action of
has proven to be a criticai problem in the oi) refining and pipeline tensile stress, relatively high hardness and corrosion in the presence
industries. Sour corrosion can take three forms-pitting, cracking, and of water and H2S (Fig. 5.35). SSC results from absorption of atomic
blistering-eventually causing a leak or rupture of the component. hydrogen that is produced by the sulfide corrosion process on the met-
Wet H 2S corrosion became a priority concern in 1984, following a mas- al surface. Some hydrogen diffuses into the steel and causes cracking
sive explosion and fire that killed 17 and caused extensive damage to if the steel is too hard (hardness above Rockwell C 22). The diffusion
a refinery (Fig. 5.34). The accident was traced back to a carbon steel of hydrogen depends on the fluid pressure, the component thickness,
(ASTM A 516, Grade 70) absorber tower, 50-ft tall, 94-in diameter, with and the condition of the metal surface. Wet H 2S cracking can take
a l-in thick wall, used to strip H 2S from propane-butane gas. The place under any one of the following conditions.36
tower operated at approximately 100°F and 200 psi. Ilydrogen blister -
ing had caused the replacement of the second cylindrical course from • H 2 S concentration above 50 mg/L (50 ppm)
the bottom. In 1976, a Monelliner was fitted to the bottom head and • Free water pH below 4, with some dissolved H 2S
lower course. Cracks developed at the weld between courses 1 (lined) • Free water pH above 7.6, with at least 20 mg/L (20 ppm weight) dis-
and 2 (replaced), causing the leak and subsequent explosion. It was solved hydrogen cyanide (HCN) in the water with some dissolved H 2S
reported that the weld heat-affected zone (HAZ) between the replaced
course 2 and the bottom course had a hard microstructure and the • Partia} pressure of H 2S above 0.05 psia, in gas phase
parent metal had a low toughness, which exacerbated wet H 2S crack- Also, SSC is peculiar in that it is worst at ambient room tempera-
ing of the vessel shell. ture an d decreases as the temperature increases.
Figure 5.37 Blister and surface crack.22
Figure 5.35Sulfide stress crack-

ing and Brinell hardness.17
Figure 5.38 Hydrogen-induced cracking CHIC) of steel plate.
Figure 5.36 Blister and cracks in steel plate.22
5.37 Blisters and Cracks

Hydrogen blistering is the formation ofblisters in the metal, caused by
the accumulation of absorbed atomic hydrogen H that recombines to
forro molecular hydrogen H 2 at discontinuities such as nonmetallic inclu-
sions or laminations. The molecular hydrogen, too large to further di:ffuse .....___.
through the steel, accumulates and bui1ds up pressure at discontinuities O. Smm
such as plate laminations, forming blisters and cracks (Figs. 5.36 and
5.37). These cracks are referred to as hydrogen-induced cracks (HIC).
HIC cracks, generally oriented along the circumference, could even-
tually connect radially and form steplike cracks, leading to rupture or
leakage. These steplike cracks are referred to as step-oriented hydrogen
induced cracks or stress-oriented hydrogen-induced cracks (SOHIC;
Figs. 5.38 and 5.39).
In addition to cracking, the hydrogen-charged metal experiences a loss
of ductility (elongation at rupture) or toughness (e.g., Charpy V-notch
toughness) which ís called embrittlement. Figure 5.39 Step-oriented hydrogen-induced cracking <SOillC).311
The earlier weld hardness limit had provento be insufficient to pre-

vent HIC, and new recommendations had to be developed to control HAZ
hardness. The new recommendations combine controls of base metal
chemistry (carbon equivalent and microalloying), postweld heat treat-
ment, or welding procedure qualification (preheat, high heat input).39
Atomic hydrogen can be removed by a bake-out at 350°F.
5.38 High-Temperature Corrosion

There are several forros of high-temperature degradation mecha-
nisms:l.".s (a) (b)
• High-temperature oxidation. At high temperature, a dark brittle Figure 5.4 0 (a) Mild graphitization and (b) severe graphiti-
zation.25
oxide forros on the metal surface. The corrosion rate becomes signif-
icant (10 mg/cm2 in 1000 h- 40 days) at the following temperatures:
carbon steel at 1025°F; Y2Mo steel at 1050°F; 1Cr-YzMo steel at
1100°F; 2Y,Cr- 1Mo steel at 1100°F; 18Cr-8Ni stainless steel at loOO"F;
and 18Cr-8Ni stainless steel with Mo at 1650°F.
• Carburízation. A network of carbides forros above 1600°F, causing
swelling and gross cracking and loss of ductility of the metal.
• Decarburization. At high temperature, hydrogen diffuses into the
metal and combines with carbon to forro methane bubbles, specially
trapped at grain boundaries; the surface becomes "decarburized."
• Graphitization A metallurgical transforroation of carbides in C and
C-Mo steels as they convert to graphite nodules between approxi-
mately 800 and 1100°F. The metal becomes very brittle. Aluminum- Figure 5.41 Valve in 600 psi steam
killed carbon and carbon· ·molybdenum steels, such as ASTM A 106 line.
Grade B and ASTM A 206 Grade P1, respectively, tend to graph itize
when operating above 750°F, particularly in weld heat-affected zones.
Graphitization progresses with time, from mild to severe, as illus-
trated in Fig. 5.40. If detected in the early stages, mild graphitiza-
tion can be rehabilitated by heat treatment. In the later stages the as P, Mn, Si, Sb, and Sn at the g:rain boundaries, very narrowly (a
component may have to be repaired by gouging, heat treatment, and couple of atomic layers). Solutions to temper embrittlement include40
weld deposition, or replacement of the C and C-Mo steel by a low • Improved steel making to reduce elements such as P, Mn, and Si
alloy. Spheroidization between 850 and 1400°F is similar to graphi- • Prewarming through 650 to 1000°F before applying the operating
tization in that the carbide phases evolve from a platelike forro to a load
spherical form, causing loss of strength and loss of creep resistance. • Achieving a refmed grain microstructure
• Graphitic corrosion. The deterioration of metallic constituents in • Reversing temper embrittlement by heating above 1000°F
gray cast iron, Ieaving the graphitic particles intact. • Selecting a different material
• Temper embrittlement. The loss of toughness (embrittlement) of fer- • Liquid metal embrittlement. The loss of mechanical properties when
ritic steels, primarily 2.25Cr-1Mo low-alloy steels, operating a molten metal contacts another metal. For example, a valve operated in
between 650 and 1000°F. Auger electron spectroscope (AES) analysis a superheated steam system at 600 p si and 752°F (Fig. 5.41). The
oftemper embrittled steels shows segregation of t ramp elements such valve body-bonnet bolts were A-197-B? with A-194-2H nuts (carbon
194 Chapter Five Degradatlon 195
• Corrosion coupons are machined, cleaned, and immersed in the test

solution, following standard procedures.
• The liquid test solution, and the vapor atmosphere at its surface, is
selected to closely duplicate the service conditions.
• An attempt is made at duplicating flow conditions, particularly where
erosion is involved or to prevent the accumulation of corrosion prod-
ucts that could saturate the test solution, creating passivating ftlms
that do not occur in actual service.
• Typical corrosion tests are conducted over periods of 1 to 10 days. For
low corrosion rates, the test duration, in hours, should be approxi-
Figure 5.42 Fractured bolts. mately 2000/CR where CR is the corrosion rate in mils per year (mpy).
For example, ifCR = 20 mpy, then the test duration should be 2000/20
= 100 hours.
• The test specimens are cleaned and weighed.
steel). The bolts failed in service causing release of 750oF steam (Fig.
5.42). Thc failure investigation unveiled the fact that the nuts were The most representaLive curru~:>iun Le:sl i::; achi.eved by exposing the
cadmium plated. Cadmium melts at 610°F. The melted cadmium test specimen to the actual process fluid, in the field. 11•44 Corrosion test-
caused liquid metal embrittlement of the stressed carbon steel bolt: ing in the field is typically conducted on specimens (corrosion coupons)
the molten metal reduces the strength and ductility of the grain that are either immersed in the fluid through access fittings or installed
boundaries of the exposed stressed steel. Molten metal was found in in bypass lines. The corrosion specimen, inserted through the access fit-
the fracture, and mechanical testing of specimens showed a loss of ting, can be flush with the equipment inner diameter, or protrude into
strength. the flow stream. Corrosion can be monitored continuously online, or
intermittently by retrieving the corrosion coupon.
5.39 Measuring Corrosion and
• Coupon immersion testing measures the loss of weight of a coupon
Corrosion Rates immersed in the actual fluid or a representative environment.
Corrosion testing is conducted in the laboratory or in the field. The • Electrical resistance monitoring (ER) measures the loss of metal by
onset of corrosion and corrosion rates can be calculated by direct exam- its increased resistance to the flow of current. A difficulty a1ises when
ination and weighing of the corrosion coupon, or by the measurement of conductive corrosion products deposit on the surface.
currents or potentials due to the electrochemical reaction. For uniform
corrosion, when directly measuring weight loss of corrosion coupons, • In conductive fluids, linear polarization resistance (LPR) measures
the corrosion rate is calculated as41,4 2 current changes when a small voltage is applied to sensor electrodes.
A similar technique is potentiodynamic polarization. Electrochemical
CR = M X 534.57 impedance spectroscopy (EIS) is an alternate current equivalent of
Axtxp LPR, suitable for low-conductivity environments.
• Zero resistance ammetry (ZRA) measures galvanic currents between
where CR = corrosion rate, mpy different materiais.
M = massloss, mg • Electrical field signature methods rely on measming the distribu-
A= area, in2 tion of electric potential emanating from a buried pipeline subject to
t = exposure time, h an induced current.
p = metal density, g/cm3
• Chemical analysis techniques rely on sampling and chemical analysis
Corrosion t~sting in the laboratory involves the following general ofthe process fluid to characterize its composition or detect corrosion
steps:43 products.
5.40 Coating and vinyl ester systems where operating temperature permits, with
sacrificial anodes to compensate for loss of coating and inevitable
Electrochemical corrosion takes place when a cell is formed: anode- holidays at internai surface irregularities.
cathode-conductor-electrolyte. Coating is a barrier used to prevent the
formation of such a cell, by isolating the equipment surface (which by
itself constitutes the anode-cathode-conductor) from the process fluid, 5.42 Selection
the environment, or, for pipelines and buried tanks, the soil (which con- Equipment, piping, a.nd pipeline coatings can be classified as thin film
stitutes the electrolyte). (less than 10 mils), thick film ( 10 to 30 mils), and extra thick film (over
If coating is lost over a small a.rea, forming what is called a "holida.y'' 30 mils). Where corrosion rates are expected to exceed 5 mpy, extra
in the coa.ting, or if coating loses its adhesion to the metal, letting humid- thick films are typically used. See Table 5.1.
ity, dirt, or fluids migrate between the coating a.nd the metal, then corro-
sion will proceed a.t that spot. To prevent this local corrosion, a second
TABLE 5.1 Common Coatings
line of defense may be provided in the form of cathodic protection.
Cathodic protection, by impressed current or sacrificial a.node is a back- Type Advantages Lirnitations
up to protect the loss of electrochemical shielding provided by coating.
For cathodic protection to function, current (positive charge) must be Thin Filrn <less than 10 mils)
able to flow to the pipe or equipment wall in case of coating failure or Epuxy See Thick Film.
disbondment. This causes two dilemmas when using coating in combina- Low build, severa] coats. 150°F
Vinyl Resistance to water, alkalis, salts,
tion with cathodic protection: brines, acids. Applied as low as maxirnurn. Poor solvent resistance.
40°F. Flexible, durable. Cannot be steamed clean.
• High resistivity (electrically sbielding the metal) is desirable if the
Phenolic Bond. Resistance to organic and Poor alkali resistance. Brittle.
coating is intact, but not i f it is breached. inorganic acids. solvents, Requires bak.ing 400°F.
• High cohesion (the coating does not tear) is desirable for a well- salts, water.
adhered coating, but not for a loose coating. Inorganic Cathodic protection. Applied as Organic topcoat i f water service.
zinc low as 40°F. pH 5 to 10.
The right coating, properly applied, possibly combined with cathodic Urethane- Abrasion resistance. Good Short pot life. Low filin bui1d. Not
protection, is commonly used to prevent or at least significantly reduce chemical resistance. as resistant as epoxy resins.
corrosion in piping, pipelines, vessels, and storage tanks.
Thick Film (10 to 30 mils)
Epoxy Good film forming. Self-priming. Applied above 50°F.

5.41 Common Coatings Chemical resistance. Range of
• Pipelines used to rely primarily on hot-applied coal tar enamel, and resins and curing agents. Long
shelf and pot life. Does not shield
now have shifted to fusion-bonded epoxy (FBE) or multilayer coatings cathodic protection current.
with cathodic protection (impressed current onshore and sacrificial Economical. QQod film build. Short shelf and pot life. Safety and
Coai tar
anodes offshore). Resistance to acids, alkalis, salts. hygiene limitations.
• Underground storage tanks used to rely on a 2- to 3-mil-thin asphalt
Extra Thick Film Cover 30 mils)
coating, and now these ta.nks are either being fabricated in glass fiber
reinforced plastics, or, if metallic, are covered with over 30 mils of Epoxy See Thick Film.
coating, with cathodic protection.
Polyester Good filrn build. Hard. ReslStance Up to 250°F. Not resistant to
• Small above-ground storage tanks tend to have coated bottom plates to inorganic acids, oxidizing solvents, dilute caustics. Short pot
and sacrificial anodes, and large flat-bottom storage tanks rely on agents, caustics to 50 percent. life.
their sacrificial anodes. Vinyl ester Resistance to alkalies, organic Shelf life 4 months.
acids. Flexible. Abrasion resistant.
• The intemal surfaces of tanks and vessels ca.n also be lined with 15 to Pot life 3 hours.
30 mils of fiberglass reinforced polyester and epo:xy, epoxy-phenolic,
5.43 Surface Preparation In any case, surface preparation depends on the type of coating used,
and should closely follow the supplier's requirements. The following
Coating suppliers stress that surface preparation is the most criticai step steps may be involved:
in the coating operation. There are two key objectives when preparing
a surface for coating: • Solvent clean the pipe or equipment surface to SSPC SPl.
• To remove contaminants (dirt, oil, grease, rust, soluble salts, etc.): the • Blast clean (abrasive cleaning), or pressure wash, or wire-brush
cleaner, the better. clean with a hand tool (SSPC SP2) or power tool (SSPC SP3), to the
supplier's specified profile and anchor pattern (typically with a profile
• To provide the optimum profile for adhesion. In this case, smoother on the order of 2- to 4-mils surface finish).
is not better. The optimum profile (roughness, pattern) is determined
by the coating supplier on a case basis. • Prepare surface (cleanliness and profile) to coat supplier specifica-
tions, such as SSPC SP 10 near white.
Cleaning the surface before coating cathodically protected pipe is • An acid wash (5<K wt. phosphoric acid) and 2500 psi deionized water
essential for this reason: the cathodic protection cun-ent causes the jet rinse may follow ifthere is a risk that the pipe is contaminated with
electrolysis of water and salts trapped under the coating; this in turn salt ions.64 Possibly reblast cleaning after acid wash.
will cause the coat to disbond. • Weld crowns and irregularities may have to be hand filed.
Standards for abrasive cleaning and surface prepar::~t.ion ::lrP puhlished
by the Society for Protective Coatings <SSPC), which also provides pic- • The adjacent mill-applied coating that will be covered by the field work
torial standards of cleaned steel surfaces. The degree of cleanliness is is manually abraded.
expressed in decreasing order:45-62 • The surface ts preheated l0°F above the dew point to eliminate mois-
ture (sweating).
• SSPC SP 5 (white metal, NACE No.l)
• A primer may be applied and cured, if specified by the supplier.
• SSPC SP 10 (near white)
• SSPC SP 6 (commercial, NACE No.3) When storage tanks' internai linings fail, they can be removed by
sandblasting. Alternatively, the lining is heated with an induction coil,
• SSPC SP 7 (brush off) breaking the fiberglass-steel bond, and then it can be scraped away.
A blast-cleaned surface still contains microscopic contaminants (dirt,
dust, rust, etc.). For fusion-bonded epoxy (FBE) coating, a generally 5.44 Wrap Tape
acceptable level of microscopic contamination is around 30% maxi- A wrap tape coat is usually a synthetic fabric carrier impregnated with
mum. In-mill controls can lower actual dirt contamination to less than petrolatum or siliceous fibers, woven around the pipe in a spiral manner.
lOt;r .63 Another type of contamination is by soluble salts (chlorides and Irregular shapes such as flanges may be filled with mastic before being
sulfates) that accumulate on the pipe surface during storage, shipment, covered. The wrap tape is applied over an anticorrosive primer that
and construction, reducing coating adhesion. Elimination of these salts also doubles as an adhesive. An overwrap may be used to provide resis-
is accomplished by acid wash. Because the acid wash leads to a smoother tance against impact, wear, solar heat, and electrical currents. Tape
profile, the surface may have to be blast cleaned again to achieve a coats have two advantages:
rougher finish after the acid wash.
The surface profile (roughness) ís measured as peak-to-valley height • They are easy to apply, even on marginally clean surfaces and in
of surface finish (expressed in mils), and is controlled by the choice of freezing or desert environments.
abrasive (type, size, hardness), nozzle shape, and pressure (typically • They are chemically resistant.
on the order of 100 psi).45 Profiles vary between 0.5 and 6.0 mils.
Abrasion ofthe surface can also be accomplished by water jetting with But, if not properly installed, they can Jose adhesion with time, par-
abrasives, from low pressure Oess than 5000 psi) to ultrahigh pressure ticularly under soil stresses (Fig. 5.43), and, once disbanded, they are
(over 25,000 psi). susceptible to shielding cathodic protection cun-ents.
Figure 5.43 Loss of adhesion and coating damage.
5.45 Epoxy
Figure 5.44 Application offusion-
Epoxy is a resin formed by combining epichlorohydrin and bisphenols. A bonded epoxy powder.
curing agent, an amine or polyamide, is added to convert it to a plastic-
like solid.s5 Since the 1970s fusion-bonded epoxy has been a coating of
choice in line pipe mills. Fusion-bonded epoxy is favored in mill applica-
tions beca use of its excellent adhesion, its good mechanical properties
and chemical stability, its minimal cathodic disbondment, and the fact heated to close to 450°F by high-frequency induction coils, and the pipe
that it does not shield the cathodic protection current. A sequence of surface temperature verified with temperature-indicating crayons.
operations for míll application of fusion-bonded epoxy is66 The heater is removed and the powder application machine is immedi-
ately placed on the joint to apply the epoxy powder. The thickness
• Preheat the pipe. achieved will be from 25 mils minimum to the maximum specified by
• Blast clean and inspect. the purchaser, and it will overlap with the mill-applied coat over close
to 2 in.
• Phosphoric acid wash (5% wt. ).
For practical reasons, field coating is often accomplished by brush,
• Rinse with demineralized water (to eliminate ionic contaminants roller (Fig. 5.45), or spray of two-component liqmd epoxy. Salient fea-
such as salts accumulated when shipping the line pipe from over- tures of the spray form are
seas).
• High-pressure water clean (2500 psi). • Compounds. High-solids (no volatile organic compounds) epoxy with
an amine curing agent, solvent (verify flammability}, thinner if envi-
• Heat and apply powder 8 to 30 mils (Fig. 5.44). ronment is hot.
• Cool and quench. • Performance objectiues. Chemical resistance, temperature (250°F con-
• Electrically inspect. tinuous, 400°F for short periods), thickness (6 mil per coat), color,
drying time (10 h at 70°F}, curing time (7 days at 70°F ).
Epoxy can also be applied in the field by heating on-pipe powder67 or
• Life. Shelf life (2 yr at 70°F}, pot life (6 h at 70°F).
by rolling, brushing, or spraying. In the case of powder application
(fusion-bonded) the pipe is cleaned to near-white CSSPC SP 10), pre- • Weight. 15 lb/gal.
Figure 5.45 Brush-applied epoxy

coating (Denso).
Figure 5.46 Installatíon of heat·shrink slccve.
Fiberglass-reinforced coai tar enamel has been used over heat-curable

epoxy primers for temperatur es up to 230°F.
• Plzysical characteristics. Hardness, thermal shock (5 cycles -70°F to
+ 200°F), abrasion resistance.
• Primer. Required if the steel is not cleaned to white metal, and the 5.47 Heat-Shrinkable Sleeves
finish coat is below 8 mils. Heat-shrinkable sleeves consistof cross-linked and stretched polyolefin
• Surface preparation. Remove sharp edges; clean (solvents, alkaline (polyethylene or poiypropyiene) with an adhesive (Fig. 5.46). They are
solution, steam, hot water with detergent; b1ast) surface to SSPC-SP5 commonly used in field coating of girth welds or bare spots.69 The design
or NACE No.l white metal (100 psi blast nozzle with natural abrasive is meant to resist externai forces, but there have been cases ofwrinkling
or steel grit); anchor Ctooth) pattern 20 percent of the total film thick- from soil stress.
ness; vacuum clean; heat metal l 0°F above dew point; no visible con-
densation or rust. 5.48 Multilayer Coating
• Application. Spray gun; clean 50 psi air regulated to 5 psi at nozzle, Today there are many multilayer high-performance coatings; they
8- to 12-in-wide spray; continuous agitation during spray; brush include an epoxy primer, an intermediate adhesive (because epoxy does
touch-ups. not adhere with polyolefins), anda polyolefin (such as high-density poly-
ethylene) or cross-linked backing on the outside. Other combinations
5.46 Coai Tar Enamel inciude the following.
Asphaltic bitumen is a dark viscous liquid, derivative of crude hydro- • Sprayed epoxy-based polymer concrete over FBE
carbon, with good waterproofing and adhesive pr operties. Coai tar • High-density polyethylene ( 120 to 160 miis HDPE ) extruded over a
enamei has been used to protect pipes and pipelines since th e early sprayed poiyurethane foam layer
1900s. Environmental regulations introduced in th e 1970s and 1980s
caused the closing of many coa] tar application shops, and the search • Epoxy-urethane or poiyurethane
for an aiternative coating led in most cases to fusion-bonded epoxy. • Epoxy-adhesive-polyethylene CFigs. 5.47 and 5.48)
Figure 5.47 Mill applied polyethylene coating 7 0 Figure 5.49 Ceramic-fitled composite coating.71
• Solids contents. Preferred no volatiles, 100 percent solids

• Flammable solvents. Preferred nane
• Application. Mill, field, brush, roll, spray, furnace, and so on
• Thickness. ASTM G 12. Ballpointlike magnetic pull-off gauge
• Swface preparation. For example, near white metal SSPC SP 10 and
acid wash
• Blast profile. For example, 2 mils average
• Application temperature. Rigid polyurethane may be -40 to 150°F,
spray epoxy may be 50°F min
• Substrate temperature. For example, 5 to l0°F over dew point
• Spray pressure. Where applicable may be 2000 psi
Figure 5.48 Internai polyethylene lining. 70 • Number of coats. One, two, or more
• Dry to touch. Minutes or hours
• Epoxy-coal tar enamel-fiberglass-wrap • Dry to handle. Hours or days
• Liquid coating or tape, fiberglass cloth with water-activated resin • Holiday testing. Hours or days after drying
• Ceramic-filled epoxies (Fig. 5.49) • Backfill. Hours or days
• Ultimate cure. May be days, weeks
5.49 Coating Performance
The performance objectives of coatings include the following. 72 5.50 Coating Quality Control
• Coating product. Coal tar enamel, epoxy, polyurethane, and so on Quality contrai tests for coating are as follows. 73-86
Figure 5.50 Cutout of coating for disbondment Figure 5.51 Attempt at peel-off of culouts.H7
testing.87
• Adhesion. ASTM D 4541 (e.g., in the range of 500 to 2000 psi, portable
units)
• Abrasion resistance. ASTM D 4060 (e.g., on the order of 10 to 150 mg
loss)
• Cathodic disbondment. ASTM G 95 (zero to a few mm radius dis-
bonded or blistering; Figs. 5.50 and 5.51)
• Chemical resistance. Pass or not
• Dielectric strength. ASTM D 149 (in the range of a few hundred
V/mil)
• Elongatwn. ASTM D 638 (A couple percent to as much as 60 per-
cent) Figure 5.52 Lap shear test speci-
men after failure.s•
• Flexibility. ASTM D 522 (pass or fail180° around mandrel)
• Hardness. ASTM D 2240 (range of 70 Shore D)
• Impact reslstance. ASTM G 14 (in the range of 50 to 150 in·lb)
• Penetration resistance. ASTM G 17 (5 to 15 percent) 5.51 Comparison
• Salt spray. ASTM B 117 (pass or not) Table 5.2 shows highlights from a valuable comparison of the perfor-
• Stability wet. ASTM D 870 (on the order of -30 to +150°F) mance of four types of coatings. These and other coatings were being
studied as part of a project to rehabilitate a 62-mi-long pipeline in
• Volume resistiuity. ASTM D 257 (on the order of 10 14 Ü·cm) Northern China.sa
• Water absorption. ASTM D 570 (a fraction to a couple percent)
• Water vapor permeability: ASTM D 1653 (on the order of 5 to 50 g/m 2 5.52 Practical Challenges
in 24 h>
There are severa} challenges in selecting and applying a coating sys-
• Lap shear test: ASTM D 903 and D 905 (Fig. 5.52 ) tem, either on new pipeline and equipment or for rehabilitation.s9
TABLE 5.2 Performance Test of Coatlngss 7. NACE TM0169, Standard Test Method. Laboratory Corro.~ion Testing of Metais,
NACE, Houston.
Fusion- 8. Bradford, S. A., Corrosion Control, Casti Publishing Co., Edmonton, Alberta,
Coal-tar Solid bonded Canada.
epoxy epoxy Polyurethane epoxy 9. NACE Corrosion Survey Database, NACE International, Houston, TX.
Application• B,R,S B,S s S,HC 10. NRC lnformation Notice Rupture tn Extraction Steam Pipíng as a Result of Fiow-
Accelerated Corrosion, 97-84, December 11, 1997, United States Regulatory
Thickness, mils 20 27 53 18 Commission, Washington, DC.
11. NACE RP 0497, Standard Recommended Practice Field Corrosion Evaluation URing
Surface prep. SSPC-SP10 SSPC-SPlO SSPC-SP10 SSPC-SP10 Metallic Test Specimens, NACE lnternational, Houston, TX.
Profile, mils 2 2 2 2 12. ASTM G 1, Standard Practice for Preparing, Cleaning, and Evaluating Corro.~ion
Test Specimens, ASTM International, West Conshohocken, PA.
Coats 1-2 1 1 1 13. ASTM G 46, Standard Guide for Examination and Eualuation of Pitting Corrosíon,
Dry to handle, h 12-24 3 6-8 o ASTM International, West Conshohocken, PA.
14. ASTM G 78, Standard Guide for Crevice Corrosion Testing of Iron-Base and Nickei-
Adhesion D 4541, psi 750 1850 1000 1650 Base Stainiess Alloys in Seawater and Oth~r Chloride-Containing .Aqucous
Abrasion D 4060, mg loss 160 135 40 35 Environments, ASTM International, West Conshohocken, PA.
(w.ceramic) 15. Kain, R. M., Use of Coatings to Assess the Crevice Corrosion Resistance of Stainiess
Steels in Warm Seawater. Marine Corrosion in Tropical Environments, ASTM STP
Elongation D 638, o/r 3.2 2.8 59 4.8 1399, ASTM, 2000.
16. Licina, G. J., Sourcebook for Microbiologically lnfluenced Corrosion in Nuclear
Hardness D 2240, ahoro D 65 82 RR 85 Power Plant.~. Elcctric Powcr Rcscarch Instituto, Polo Alto, CA.
Impact resistance G 14, in·lb 28 29 76 160 17. NACE, A State of lhe Art Report for Carbon Steel and Austenitic Stainless Stcel
Surfaces under Thermal lnsulation and Cementitious Fireproofing, NACE
Water absorption D 570, o/r 1.2 2 2 0.83 Publication 6H189, NACE Houston.
Vapor perm. D 1653, g!m2 12 3.8 37 7.5 18. Ashbaugh, W. G., lnspection of Vessels and Piping for Corrosion under In.sulation,
Material.~ Performance, Vol. 29, July 1990.
*B = brush, R = rollcr, S =spray, and HC =fluidized bed heat cured. 19. ASTM 880, Special Publication 880, Corrosion Under Insulation, ASTM
International, West Conshohocken, PA.
20. Moniz, B., Kobrin, G., NACE lntemational Symposilun on Process Piping,
Inspection, Maintenance, and Preuention of Corrosion of Piping and Equipmenl
• Tests are typically limited to new, freshly applied coatings. Under Thermal Insulation, 1993, NACE, Houston.
• Tests are carried out on different samples, not accounting for accumu- 21. Johnson, C J., Keep Water Flowing to Reduce the Pot€'ntial for MIC, Power, March,
1987.
lation of adverse conditions. For example, water and oxygen penetra- 22. IGefner & Assoei ates, Worthington, OH.
tion at operating temperatures are often not reflected in tests. 23. Huchlcr, L. A., Monitor microbiological populations in cooling water, Hydrocarbon
Processin!(, September, 2002.
• Tests rely on the extrapolation of 30-h to 20-yr and possibly 50-yr 24. Pope, D. H., Zintel, T. P., Methods for lnvestigating Underdeposit Microbiologically
performance. Intluenced Corrosion, Materiais Performance, 29(11), 1989.
25. Thielsch Engincering, Cranston, RL
• Tests are designed to fail the coating, not to predict its life. 26. Delp, G. A., et. ai., Erosion-Corros1on in Nuclear Power Plant Steam Piping: Causes
and lnspection Program Guidelínes, EPRI NP-3944, April, 1985.
• Coating life depends on surface preparation and envíronment. 27 API RP 14E, Recommended Practice for Design and lnstallation of Offshore
Production Platform Piping System.~. American Petroleum Institute, Washington,
DC.
References 28. USNRC, Information Notice 86-106 Feedwater Line Break, Dec. 16, 1986, and
Supplement 1, United States Regulatory Commission, Washington, DC.
1 Dillon, C. P., Forms of Corrosion Recognition and Prevention, NACE 29. USNRC, Information Notice 82-22, Faílure in Thrbine Exhau,çt Lines, Unitcd States
International, Houston. Regulatory Commission, Washington, DC.
2. Dillon, C. P., Corrosion Control in the Chemical Proces.~ lndustries, Materiais 30. USNRC, lnformation NotiC€' 92-35, Higher than Predicted Erosion-Corrosion m
Technology Institute of the Chemícal Process Industries, St. Louis, MO Unisolable Reactor Coolant Pressure Boundary Pipirt!( lnside Containment at a
3. ASTM G 15, Standard Terminolol{y Reiatmg to Corrosíon and Corrosion Testing, Boiler \Vater Reactor, May 6, 1992, United States Regulatory Commission,
ASTM International, West Conshohocken, PA. Washington, DC.
4. ASM Metais llandbook, Volume 13, Corrosion, Amcncan Society of Metais, 31. Cragnolino, G., et. al., NUREG/CR-5156, Review of Ero.~ion-Corrosion in Singie·
Materiais Park, OH. Phase Flow, April. 1988, United States Regulatory Commission, Washington, DC.
5. ASM, Handbook ofCorrosion Data, American Society ofMetals, Materiais Park, OH. 32. Welding Research Council Bulletin, WRC 382, Nuclear Piping Criteria for Advanced
6. Herro, H. M., Port, R. D., The Nalco Guide to Coolíng Water Systems Failure Light Water Reactors, Volume 1- Failure Mechanisms anti Correctiue Actions, June,
Analysis, Nalco Chemical Company, McGraw-Hill, New York. 1993, Pressure Vessel Research Council.
33. Welding Research Council Bulletin, WRC 490, Damage Mechanisms Affecting Fiud 60. SSPC VIS 1, Visual Standard for Abrasiue Blast Cleancd Steel, The Society for
Equipment in the Fossil Electric Power Industry, J. D. Dobis and D. N. French, April Protecttve Coatmgs (SSPCJ, P.ittsburgh, PA.
2004. 61. SSPC VIS 2, Standard Method of Eualuating Degree of Rusting on Painted Steel
34. ASM, Atlas of Stress Corrosion and Corrosion Fatigue Curves, American Society of Surfaces, The Society for Protective Coatings cSSPCJ, Pittsburgh, PA.
Metais, Materiais Park, OH. 62. SSPC VIS 3, Visual Standard for Power- and Hand-Tool Cleaned Steel (Standard
35. ASTM A 262, Standard Practicex for Detecting Susceptibility to Intergranular Attack Reference Photographsl, The Society for Protective Coatings <SSPC), Pittsburgh,
in Austenitic Stainless Steels, ASTM International, West Conshohocken, PA. PA.
36. NACE RP 0296, Guidelines for Detection, Repair, and Mitigatíon of Cracking of 63. Neal, D., Good Pipe Coating Starts with Properly Prepared Steel Surface, Pipe Line
Existing Petroleum Refinery Pressure Vessels in Wet H2S Enuironments. & Gas lndustry, March, 1999.
37. Bullen, M. L., et. ai., Refinery Experience with Cracking of Pressure Vessels Exposed 64. Weldon, D. G., et. ai.. The Effect of Oil, Grease and Salts on Coating Performance,
to Wet H~ Enuironments, Pressure Vessel Research Council, Managing lntegrity of Journal of Protective Coatings & Linings, June, 1987.
Equipment in Wet H2S Service, October 10-12, 2001, Houston. 65. Petrie, E. M., Handbook ofAdhesives and Sealants, McGraw-Hill, New York.
38. Courdeuse, L., et. ai.. Carbon A1anganese Steels for Sour Service- Improvement of 66. Kazemi, l\1. A., Nose, B. Y.. Fusion bonded epoxy pipe coating: Preparation and appli-
HIC and SSC Resistance, Pressure Ve:;sel Research Council, :\1anaging Integrity of cation make a big difference, Journal of Protectit•e Coatings and Linings, May 1992.
Equ1pment in Wet H2S Sen;ce, October 10-12, 2001, Houston. 67. NACE RP0402, Field-Applied Fusion-Bonded Epoxy fFBEI Pipe Coating Systems for
39. NACE RP 0472, Methods and Controls to Preveni In-Service E1wironmenta/ Girth Weld Joints: Application, Performance, and Qualit_v Contrai, NACE
Cracking ofCarbon Steel Weldments in Corrosive Petroleum Refining Enuironments- International, Houston.
Item No. 21006. NACE Intemational, Houston. 68. Denso North America, Houston.
40. Viswanathan, R., Damage Mechanisms and Life Assessment of High-Temperature 69. NACE RP 0303, Standard Recommended Practice Field-Applied Heat-Shrinkable
Components, ASM lntcrnational, Metais Park, OH. Sleeues for Pipelil!es: Application, Performame, anel Quality Control-Item No.
41. Roberge, P. R., Handbook of Corrosion Engineering, McGraw-Hill, New York. 2ll01, NACE International, Houston.
42. ASTM G 31, Standard Practice for Laboratory lmm~rsiun Cuuvsivn Thsting of 70. Protubo, Rio de Janeiro, Brazil.
Metais, ASTM International, West Conshohocken, PA. 71. Chesterton, ARC Composites for Abrasion, A. W. Chesterton Co., Stoneham, MA.
43. NACE TM0169, Standard Test Method. Laboratory Corrosion Thsting of Metals, 72. Guan, S. W., et. ai., Pipeline Rehabilitation ai Ali Environmental Temperatures with
NACE International, Houston. Advanced 100'lf Solids Structural and Rigid Polyurethane Coatings Technology,
44. ASTM G 4. Standard Guide for Conducting Corrosion Thsts in Field Applicatinns, Pipeline Rehabilitation and Maintenance, October 2003, Berlin.
ASTM lnternational, West Conshohocken, PA. 73. ASTM B 117, Standard Practice for Operating Salt Spray (Fog) Apparatus, ASTM
45. SSPC AB 1. Mineral and Slag Abrasit•es, The Society for Protective Coatings CSSPCJ, International, West Conshohocken, PA.
Pittsburgh. PA. 74 ASTM D 149, Standard Test Method for Dielectric Breakdown Voltage and Dielectric
46. SSPC COM, SSPC Surface Preparation Specifications, Surface Preparation Strength of Solid Electrical lnsulating Materiais at Commercial Power Frequencies,
Commentary for Steel and Substrates, The Society for Protective Coatings <SSPCJ, ASTM Intemational. West Conshohocken, PA.
Pittsburgh, PA. 75. ASTM D 257, Stondard Test Methods for DC Resistance or Conductance of
47. SSPC SP 1, Soluent Cleaning, The Society for Protective Coatings (SSPCJ, lnsulating Materiais, ASTM International, West Conshohocken, PA.
Pittsburgh, PA. 76. ASTM D 522, Standard Test Methods for Mandrel Bend Test of Attached Organic
48. SSPC SP 2, Hand 'lbol Cleaning, The Society for Protectivc Coatings (SSPCJ, Coatings, ASTM International, West Consbohocken, PA.
Pittsburgh, PA. 77. ASTM D 570, Standard Test Method for Water Absorption of Plastics, ASTM
49. SSPC SP 3, Power 'lbol Cleaning, The Society for Protectivc Coatings CSSPCJ, International, West Conshohocken, PA.
Pittsburgh, PA. 78. ASTM D 870, Standard Practice for Testing Watcr Resi.~tance of Coatings Using
50. SSPC SP 5, White Metal Bla~t Cleanwg, NACE No.l. The Society for Protective Water Immersion, AST~1lnternational, West Com;hohocken, PA.
Coatings cSSPCl, Pittsburgh, PA. 79. ASTM D 1653. Standard Test Methods for Water Vapor 7}-ansmission of Organic
51. SSPC SP 6. Commercial Blast Cleaning, NACE No. 3, The Society for Protective Coating Films, AST.M International, West Conshohocken, PA.
Coatings cSSPCl, Pittsburgh. PA. 80. ASTM D 2240, Standard Test Method for Rubbcr Property-Durometer Hardness,
52 SSPC' SP 7. Joint Surface Preparation Standard Brush-Off Bla.~l Cleaning, NACE ASTM Intemational, We:st Conshohocken, PA.
No. 4, The Society for Protectivc Coatings cSSPCJ. P.ittsburgh, PA. 81. ASTM D 4060, Starzdard Test Method for Abrasion Resistance of Organic Coatings
53. SSPC SP 8, Pickling, The Society for· Protective Coatings <SSPCJ, Pittsburgh, PA. by the Taber Abraser, ASTM International, West Conshohocken, PA.
54. SSPC SP 10, Near-White Metal Blast Cleaning, NACE No. 2, The Society for 82. ASTM D 4541, Standard Test Method for Pull-Of( Strength of Coatíngs Usmg
Protective Coatings CSSPCl, Pittsburgh, PA. Portable Adhesion TI!~Sfl!rs, ASTM International, Wcst Conshohocken, PA.
55. SSPC SP 11, Power Tool CleanÍil/i to Bare Metal, The Society for Protective Coatings 83. ASTM G 12, Standard Test Method for Nondestructiue Measurement of Film
cSSPCJ, Pittsburgh, PA. Thickness of Pipeline Coatings on Steel, ASTM lntcrnational, West Conshohocken,
56 SSPC SP 12. Surface Preparation and Cleaning of Metals by Waterjetting Prior to PA.
Recoating, NACE No. 5, The Socicty for Protective Coatings cSSPCJ, Pittsburgh, PA. 84. ASTM G 14, Standard Test Method for Impact Resistance of Pipelme Coatmgs
57 SSPC PC 13, Surface Preparation of Concrete, NACE No. 6, The Society for rFalling Weight TestJ, ASTM International, West Conshohocken, PA.
Protecti\·e Coatings cSSPCl, Pittsburgh. PA. 85. ASTM G 17, Standard Test Method for Penetration Resista11Ce of Pipeline Coatings
58. SSPC SP 14, Industrial Blast Cleaning, NACE No. 8, The Society for Protective (Blunt Rod), ASTM International. West Conshohocken, PA.
Coatings cSSPCJ, Pittsburgh, PA. 86. ASTh1 G 95, Sta11dard Thst Method for Cathodic Disbondment Test of Pipeline
59. SSPC SP 15. Commercial Grade Pou·er Tool Cleaning, The Soc1ety for Protective Coatings, ASTM International, West Conshohocken, PA.
Coatings cSSPCJ, Pittsburgh, PA. 87. Citadel Technologies, 1\Jlsa, OK.
212 Chapter Five
Chapter
6
88. Guan, S. W., et. al., Rigid-PU Coating Addresses Chronic Corrosion on Northern
Chinese Gas Pipeline, Oil & Gas Journal, October 4, 2004.
89. Dabiri, M., et. al., The Use of Coatings in Rehabilitating North America's Aging
Infrastructure, Pipeline & Gas Journal, February 2001.
90. White, R. A., Materials Selection for Petroleum Refineries and Gathering Facilities,
NACE lntemational, 1998.
lnspection
6.1 Principies of lnspection

The ASME code refers to examination as a nondestructive examina-
tion activity (e.g., making a radiography is an examination function),
whereas inspections refer to the oversight quality assurance (e.g., the
owner inspects the job to make sure that code and contractual require-
ments have been met). In this book, inspection refers to the examina-
tion of equipment that has been in service. These examinations are
conducted to determine the equipment's condition and its fitness for
continued service. Inspections are primarily nondestructive. When plan-
ning inspections, the key questions are: Why? What? Where? How?
When?
6.2 Why?
What will be the purpose ofinspection? There are normally three reasons
to inspect:
• Safety. Worker and public safety, and environmental protection
• Production. Equipment reliability, failure prevention, planning for
repairs or replacements
• Regulation. Regulatory compliance
In reality, regulations and safety are one and the same, but, in certain
cases, the operating company may feel that inspections are carried out
simply to satisfy regulations. But, as regulators and operating com-
panies move towards risk-based inspections, this disconnect tends to
disappear.
213
214 Chapter Stx lnspectton 215
6.3 What?
What structure, system, or component should be inspected? In our case,
what pipe systems, pipeline sections, vessels, and tanks should be
inspected? Selection of equipment is best achieved through a serious
look at the risk offailure: the likelihood offailure and the consequence
of failure. This is the principie of "risk-based inspection" (RBI), also
referred to as "risk-infonned inspection," which will be addressed in this
chapter.
6.4 Where?
Once the equipment is selected for inspection, where should it be
inspected? This is possibly the most difficult question of all.
Ideally, at this stage plant staff has defined the potential degrada-
tion mechanisms, and inspections can focus on those locations where
degradation mechanisms are believed to be at play, for example,
Figure 6.1 Remote visual ínspection by borescope shows
crack.
• Cracking due to thermal cycling at nozzle injection points
• Erosion in wet steam service at elbows and tees
6.6 When?
• Corrosion at the liquid-vapor interface
Setting inspection intervals is an outcome of the fitness-for-service
The selection of inspection points is in itself a diffi.cult task, but the analysis of prior inspection results. Through inspections we determine
difficulty is compounded when we note that failures have occurred only degradation mechanisms and rates of degradation. By extrapolation we
feet away from previously inspected and apparently sound material. predict end of life, and finally the inspection interval. A good rule of
This is particularly true with cracking and localized corrosion. Ideally thumb is to start with a ''half remaining life" approach. For example, if
a thorough internai and externai visual inspection ofthe equipment or inspections indicate a corrosion rate that will cause leakage or failure in
pipe segment would be conducted to pinpoint areas of degradation and six years, the inspection interval would be set at half the remaining life,
guide volumetric inspections. or three years.
For example, the crack in Fig. 6.1 was discovered by visual borescope An alternative to the half remaining life approach is to apply statisti-
inspection of vessel internais; the rest of the vessel appeared to be in cally based inspection methods commonly used for rotating machinery. 1
good condition. This permits us to focus ultrasonic examination on this The first step would be to gather data on mean time between failures
area. This challenge does not apply to oil and gas pipelines which are (MTBF),
fully inspected volumetrically by means of intelligent pigs (inline
MTBF = M~R
inspection tools), as described later.
6.5 How? where MTBF = mean time between failure, months

Knowing why and where to inspect, and knowing what we are looking P = population
for, we select the inspection technique that has the best probability of MFR = monthly failure rate, 1/month
detection (POD) and accuracy. An SNT-TC-lAlevel UI certified inspec-
tor should assist in the selection of the right inspection technique, and The MTBF can be expressed as a reliability R,
inspection should be carried out by a certified person. Common inspec- 1
tion techniques are described later in this chapter. R= xlOO
exp (t/MTBF)
216 Chapter Slx lnspectlon 217
where R = reliability function, % and t = operating time, months. ranked on a Hst from the highest risk number to the lowest risk number.
Conversely, the inspection interval tinsp necessary to achieve a capture Equipment at the top of the list will be most often and most thoroughly
rate CR (probability of detecting failures before they occur) is inspected, and equipment at the bottom of the list will be least often
inspected, or may simply be run to failure i f the safety and cost conse-
t;n"P == I MTBF x ln (CR) I quences of the failure are acceptable.
Risk can also be represented as a matrix of (likelihood of failure X
where t ms = inspection interval, months, and CR == capture rate. consequence of failure ). The matrix is then divided into zones of high,
The third and most comprehensive method for setting inspection medium, and low risk. This matrix, rather than the single risk number,
intervals is a risk-based approach where we not only consider likelihood is the approach selected by the American Petroleum Institute (API) in
offailure but also consequence. In this chapter, we first review the con- its standards API 580 and 581. API 580 is a general introduction to the
cepts of risk-based inspection (RBI), which will help us determine what concepts of risk and risk-based inspections. API 581 is the detailed
to inspect, and the inspection tools that will help us determine how to implementation procedure.
inspect. The question of where to inspect is addressed in the chapter Figure 6.2 illustrates a risk matrix. This is a 5 X 5 matrix, where
on degradation mechanisms. likelihood of failure is the vertical axis (VL = very low, L = low, M =
medium, H = high, and VH = very high), and consequence of failure
6.7 Risk-Based lnspection-What ls Risk? is the horizontal axis, also labeled VL to VH.
Equipment and subsystems in the VH-VH box have the highest risk
RBI is the process ofprioritizing and conducting equipment inspections and deserve the highest inspection priority. In the same manner,
on the basis ofthe likelihood (probability) offailure and the consequence VL-VL is the lowest risk.
o f this failure. The combination of probability and consequence of fail- The matrix can, in turn, be subdivided into regions of high, medium,
ure is what defines "risk.'' and low risk, as indicated in Fig. 6.2. The boundaries among high, medi-
RBI is a means ofprioritizing inspections to optimize production (by um, and low risk are defined by the operating company, and-in some
minimizing risk of unscheduled shutdowns and repairs) and safety cases- by an insurance company or the regulator. These boundaries are
(by minimizing the risk of leakage or rupture, and their consequence logical but subjective, and a particular facility may decide to alter the
on workers, the public, and the environment). RBI is also a tool to help boundaries. For example, one company may decide to extend the low
prioritize inspection budgets and schedules. risk zones to include ali ofthe "Consequence VL" column, as is done in
Inasmuch as we are interested in static equipment, the RBI list will Fig. 6.2.
encompass equipment (tanks and vessels) and subsystems (portions
of piping systems or pipelines). Components (such as va~ves) may ?e
treated as part ofthe piping subsystem or separately as a ptece of eqw~ I
ment. Active mechanical equipment (such as pumps and compressors) ts VH

HIGH RISK
not addressed here. f---
RBI of static equipment is a tool commonly used in the nuclear power
H
industry, under the name Risk-Informed-Inspections. It is also applied o
and imposed by federal and state regulation for oil and gas pipelines, o 1- -
o
under the name Integrity Management Plan (IMP).2 It is well under- I
M MEDIUM RISK
::i
way in most refineries and petrochemical operations, following API 580 w
~
and API 581 3.4 under the name RBI. It is applied to various degrees ::i
(depending o~ the operating company and state or local regulations) in L
the chemical process and fossil power industries.5 LOW RISK
VL
6.8 A Number or a Matrix -
VL L M H VH
Risk can be represented as a single number that accounts for both like-
CONSEQUENCE
lihood of failure and consequence of failure, in which case, each piece
Figure 6.2 Risk-based inspection matrix.
of equipment, or subsystem, can be assigned a single risk number and
218 Chapter Slx lnspection 219
TABLE 6.1 Example RBI Ranklng Plantwlde Percent Equipment inspected. Equipment in the medium risk zone may be inspected with a
VL L M H VH Totais longer interval. RBI can also guide the selection ofinspection techniques
(How), inspection intervals (When), and inspection location (Where).
VH o 1 .Q Q l 2
H 2 6 4 i Q 21
6.11 ls RBI a Cost Saving?
M o 1 3 2 2. 8
L 2 5 5 6 2 20
RBI is a logical inspection prioritization technique. It is a formal and
intelligent process that forces the best plant knowledge to be compiled,
VL 5 12 12 11 9 49
assembled, and presented in a structured manner; drawing on expertise
Totais 9 25 24 23 19 100 % in (1) inspection, (2) design, (3 ) materiais and corrosion, (4) operations,
and (5) safety.
RBI can mean higher inspection costs. If the facility operated on the
6.9 The Objective of RBI basis of corrective maintenance (run to failure) it did very little inspec-
RBI is the evolution from corrective maintenance (run to failure) and tion, then the RBI process will increase inspection costs, but it will also
reduce unplanned shutdowns, improve safety, and reduce overall oper·
preventive maintenance (inspect at flxed intervals) towards predictive
maintenance (set inspections on the basis oflikelihood offailure, based ating costs.
RBI can mean lower inspection costs. Ifthe facility implemented a pre·
on previous inspections and experience) and purposeful corrective main·
ventive inspection program (inspecting many things at fixed intervals),
tenance (many components and subsystems, those with low risk, may
still be run to fail ure, but with a known and accepted risk). then RBI will help eliminate some inspections and extend others. For
example, common preventive inspection schedules in process plants are:
The final objective of RBI is to populate the matrix with points, each
point being a piece of equipment (e.g., a tank), a component (e.g., a • Externai visual examination every three to tive years
valve). ora subsystem (e.g., a gas pipeline from compressor station X
to mile marker Y, or a piping system inside building Z). • Interna] or ultrasonic examination every flve to ten years
As an example, the matrix of Table 6.1 is the result of a plantwide • Visual externai inspections during operator rounds
application of RBI, and includes nearly 1000 pieces of equipment and
subsystems. It follows the same nomenclature as Fig. 6.2. The entries It is in plants that impleme nt these fiXed-schedule inspections that
are expressed in percent, so that the highest risk box (VH) comprises RBI is most likely to reduce inspection costs, by extending inspection
1 percent = 10 pieces of equipment or subsystems. The High Risk zone periods, and eliminating altogether inspections of little value.
(underlined values) has 12 percent = 120 pieces of equipment and sub·
systems. This is similar to risk-ranking results obtained in a chem ical 6.12 Qualitative or Quantitative RBI
complex where 1000 of 12,000 components (8 percent) ended up in the
RBI program.r> Likelihood and consequencc (risk) can be estimated in one ofthree ways:
(1) qualitatively, based on experience and expertise; (2) semiquantita-
tively, based on experience augmented by some quantitative guidelines,
6.1O Necessary and Sufficient and (3) quant itatively, based on detailed analysis of failure modes and
Inspection programs ought to be based on the obvious fact that inspec- effect s.
tions must be necessary and sufficient.
• Qualitative RBI. In a qualitative RBI, the likelihood offailure could
• They must be necessary and not inspect what does not need to be be based on plant-specific or industrywide experience, coupled with
inspected. corrosion and mechanical integrity expertise. The consequence offail-
ure could be based on plant experience, and management, process, and
• They must be sufficient and do inspect what needs to be inspected. safety expertise.
RBI, by ranking equipment risk, will help us decide what equipment • Semiquantitative RBI. The likelihood in a semiquantitative RBI would
needs to be inspected: All equipment in the high-risk zone should be be a number equal to a generic failure rate multiplied by correction
220 Chapter Six lnspection 221
factors to account for the specific plant and process environment. The • Step 4-lnspections. Schedule inspections, determine inspection loca-
consequence would be an estimate of, for example, production loss, tions and techniques, and develop acceptance criteria. Where inspec-
costs of recovery, fire or explosion hazard, worker and public safety, tions are not feasible, reduce risk by modifying operations, adding
and environmental impact. alarms, isolation, sprinklers, dikes, reducing inventory, online monitor-
• Quantitative RBI. The likelihood of failure in a quantitative RBI ing of chemistry or process parameters, adding corrosion inhibitors,
would consider the demand on the equipment (operating and abnor-
malloads) and its capacity (mechanical and structural properties,
material condition, degradation and damage), and would compare Step-1
demand to capacity in a probabilistic manner. In other words, it would Policy
determine the chances that demand exceed the actual, degraded Responsibilities
capacity ofthe system or equipment, and the likelihood ofthis occur-
ring. The consequence of failure would be based on a ful1 safety analy-
sis, considering failure modes (leak, malfunction, collapse, or rupture)
~
and effects in the plant and at its boundaries. The safety analysis Step-2
would apply techniques such as event trees, Hazop, failure modes and Methods
effects analysis, and so on. Procedures
Of course, as we progress from qualitative to quantitative RBI, the ~

results (the relative rank of equipment) are more accurate, but the cost
and the time required to achieve this accuracy could be prohibitive. In Step-3
+---
modern plants and facilities, a safety analysis may already exist, and Risk Ranking
systems may already be ranked on the basis of their consequence of
failure. In this case, the consequence score is already assigned. What
is left to do is to assign the likelihood of failure score for each piece of ~
equipment and subsystem, in its actual, degraded condition.
Step-4
lnspections
6.13 RBI: A Seven-Step Process
RBI is a seven-step process (Fig. 6.3). ~
• Step 1-Policy. State goals, objectives, schedules, responsibilities, and Step-5
charters for implementing RBI. The goal cannot be "to reduce inspec- Fitness-for-Service
tion costs," but rather "to optimize inspections, to establish necessary-
and-sufficient inspections." Also, commit to fitness-for-service as a tool
to (a) intelligently analyze inspection results, and (b) adjust inspec- ~
tions accordingly. Assign a budget to support the effort. Step-6
• Step 2-Methods and procedures. Select standards (ASME,6 API,3.4 Run-or-Repair
NBIC NB-23,7 etc.), criteria, software if any, and issue the implement- Next lnterval
ing procedure.
• Step 3-Risk ranking. Colled design records and previous inspection !
data, identcy corrosion mechanisms and mark them on P&IDs referred
to as "corrosion loops," rank likelihood, review or develop safety analy- Step-7 f---
Feedback Figure 6.3 RBI steps.
sis and determine consequences, develop risk ranking, and populate
the risk matrix.
222 Chaptar Six lnspection 223
planning for coating and lining, and implementing indirect assess- TABLE 6.4 Qualitativa Ranking of Consequence--Example 1
ment for underground commodities (cathodic protection or coating sur- Category Description
veys). Save inspection results in a standard, accessible, retrievable
format, preferably with the component maintenance history record. VH Public or worker fatality
• Step 5-Fitness-for-service. Analyze inspection results using fitness- H Worker injury

for-service techniques, not unsubstantiated judgment. Establish M Significant repair and shutdown costs
remaining life. Document the FFS analysis and decisions the same L Repair and shutdown costs
way as the inspections were documented in Step 4. VL Repair costs
• Step 6-Run-or-repair decision. FFS will have two possible outcomes:
(a) keep running, in which case state the next inspection date, or
TABLE 6.5 Qualitativa Ranking of Consaquence--Example 2
repair, or (b) repair, in which case prepare the conceptual design of the
repair, and state when the repair needs to be implemented. Category Description
• Step 7-Update the RBI program. lf necessary, revise the likelihood Major companywide emergency
VH
offailure by trending the latest measured degradation; possibly revise Major cleanup, months
the consequence of failure if process or plant conditions have changed. Widespread long-term significant effects on environment
Significant effect on community
Fatality or long-term public health impact
6.14 Qualitative RBI H Division emergency response
Significant cleanup, weeks
The likelihood of failure and the consequence of failure can be estab- Locallong-term effects on environment and small community
lished based on experience. This constitutes a fully qualitative RBI. lnjury or illness
Tables 6.2 to 6.5 provide examples of cri teria that can be used to rank M Field-level emergency response
failure likelihood and consequence. Cleanup, weeks
Moderate effects on environment and small community
Minar lost-time injury
TABLE 6.2 Qualitativa Ranking of Likalihood-Exampla 1
L Local emergency response
Category Description Cleanup, days
Minor effects on environment
VH Happens with some regularity in the facility First-aid cal"e
H Has happened once in the facility VL lnconsequentialleak
M Has happened in similar system in industry
L Has not happened in industry, but is possible
VL Has not happened, and unlikely to happen 6.15 Example of Qualitative RBI-Steam
Systems
A process facility decided to undertake a risk-based inspection of its
TABLE 6.3 Qualltatlve Ranklng of Llkellhood-Example 2
steam systems. s The approach selected was the 5 X 5 qualitative RBI
Category Description matrix. The ranking of consequences was developed by a senior group
offacility engineers, and the ranking oflikelihood and final risk rank-
VH Frequently or repeatedly (-10'7!' per year)
ing was developed by a team of steam experts.
H Severa! times or probable r-1% per year> An important consideration in assessing the consequence of a steam
M Occasional or likely (-0.1% per year) leak on worker safety isto note the following.9
L Not likely or remote r-1/100% per year)
• Condensate at 300°F (saturation pressure 50 psi) leaking from a %-in
VL Improbable or practically impossible (1/1000% per year)
bole cools to 150°F ata distance of 20 in, in 80°F atmosphere.
• A steam leak in a confined or enclosed space can be deadly. The thresh- Next is likelihood. In looking at likelihood, failure is defined as a leak
old of burn and pain is typically around 120°F. When the volume of (pinhole or small stable crack) or a rupture (large failure, e.g., over half
steam in a steam-air mixture is below 12 percent, the steam condens- the circumference or separation) of the base material or welds. It does
es only on surfaces cooler than 120°F, and the condensate is below the not apply to leaks or ruptures in mechanical joints such as flanged,
threshold of pain. But if the volume of steam exceeds 12 percent, the threaded, or grooved joints and couplings because our objective is the
condensing temperature will exceed 120°F and wiill therefore burn as inspection ofbase metal and welds. Unlike consequence, likelihood is not
it condenses. This is why a steam leak in an enclosed or confined space, governed by the single worst line ofinquiry; instead it is an average of
where the steam volume quickly exceeds 12 percent, is dangerous. each line of inquiry score. The likelihood lines of inquiry and assigned
points are:s
Following are the lines ofinquiry for consequence; the highest-ranked
consequence is the final resultant consequence score.s Assigned points:
• VL = O to 20
Effect on public:
• L = 21 to 40
• Nane= VL • M = 41 to 60
• Low = H • H = 61 to 80
• Possible = VH • VH = 81 to 100
Effect on worker health and safety: Prior failures:
• None = VL • None = VL
• Potentiallost time = L • Occurred in similar systems in industry = L
• Probable lost time or injury = M • Occurred in similar systems at the plant = M
• Potential fatality = VH • Occurred in this system = VH
Effect on environment: Potential degradation mechanisms:
• Reportable = VL • None = VL
• Alert =L • Possible = M
• Site emergency = M • Known = VH
• Area emergency = H
Novelty of process:
Effect on production loss:
• None = VL
• Less than one month = VL • Some = M
• One to three months = L • New process = VH
• Three to four months = M
• Four to six months = H Abnormalloads in service:
• Over six months = VH • None = VL
• Low possibility = L
Recovery costs do not include cost ofinvestigation, cost to change pro- • Possible = M
cedure.s or retrain personnel, any modification costs, or loss production • Anticipated = VH
cost: Age:
• Less than $0.5 M = VL • Less than 5 years = VL
• 5 to 15 years = M
• $0.5 to 1 M = L • 15 to 30 years = H
• $1 to 2M = M • Over 30 years = VH
• $2 to 5 M = H Plantwide steam lines were evaluated using the above rules for like-
• Over $5 M = VH lihood and consequence, and the results are plotted in Fig. 6.4.
TABLE 6.6 Example of Generic Fallure Rates•

VH
! HIGH RISK
Leak hole diameter
10-in pipe
3/4 in
2 X 10 ; 3
1 in
X 10 8
4in
8 X 10 8
Failure
2 " 10- 8
@ Vessel 4 10 5 1 X 10- • 1 X 10- 6 6 X 10 6
o
H
~ ê @ @ X
' Pipe per foot and year, vessel per year.4

o
o
:c
:J M MEDIUM RISK
w The correction factor CF may be a multiplier that accounts for five
~
:J
factors:
L §) @® • An environment factor EnF that accounts for weather and natural
LOW RISK @@
- -- - -
phenomena
VL
• A mechanical factor McF that accounts for mechanical complexity,
VL
I L M H
I
VH
quality of construction, operating pressure and temperature, equip-
ment age, vibration, and transients
CONSEQUENCE • A process factor PrF that accounts for maintenance record, pruce:s:s
Figure 6.4 Risk matrix steam systems. stabili ty, and overpressure protection
• A corrosion factor CoF that accounts for likelihood of corrosion, and
corrosion rate compared to the component overthickness
The symbol R stands for a rupture and L is a leak. Four categories
are plotted, all others are considered low risk: • A managcment factor MgtF that accounts for quality of operations,
engineering, inspections, and maintenance
• Category 1 (points R1 and L1) are steam lines -in OSHA Confined
Space that require an entry permit. As a departure from the technique m API 581,4 and to reflect the rei-
ative importance ofthese five factors, the total correction factor should
• Category 2 (points R2 and L2) are steam lines in enclosed spaces with give heightened importance to the corrosion factor CoF and the man-
little ventilation (offices, corridors, cafeteria, etc.). agement factor MgtF; for example,
• Category 3 (points R3 and L3) are outdoor steam lines larger than 2
in, operating above 150 psi and within 3 ft of traffic areas (walk- CF = CEnF + McF + PrF) X CoF X MgtF
ways, roads).
• Category 4 (points R4 and L4) are outdoor steam lines smaller than 6.17 Semiquantitative RBI-Consequence
2 in or operating below 150 psi but wíthin 3 ft of traffic areas. In a semiquantitative RBI, the consequence of failure is process-specific.
• Category 5 (points R5 and L5) are steam tines within 3ft of elevated In the case of a refinery, consequences include:
walkways or ladders without a cage.
• Vapor cloud explosion CVCE)
6.16 Semiquantitative RBI-Likelihood • Boiling liquid expanding vapor explosion (BLEVE)
• Pool fire
In a semiquantitative RBI, the likelihood of failure is equal to a generic
failure rate multiplied by a system or equipment-specific correction factor • Jet frre
CF. The generic likelihood of failure can be obtained from industrywide • Flash fire
reliability databases or, even better, from plant-specific data. Generic
failure rates can be obtained from API 581.4 A couple of examples are Consequences are expressed in terms of explosion area or fire damage
illustrated in Table 6.6. area, and fatalities area. In the case of chemical processes involving
1
lnspection 229
228 Chapter Slx
toxic fluids, the five categories of consequence in the steam example

above could be used for a semiquantitative RBI.
6.18 Fully Quantitative RBI-Likelihood where

In a fully quantitative risk analysis, the likelihood offailure is obtained
dg = 1-~ flt
by a statistical comparison of demand (applied stress) to capacity (remain- dS1 3 t
ing strength). This is done by first defining a limit state function. A
limit state function is a function, that we call g, equal to the difference dg 28,
between the capacity and the demand -----
d~t 3t
g = Cap- Dem dg D
where g = limit state function, psí dP =- 2t
Cap = capacity (remaining strength) of degraded material, psi
Dem = demand (applied stress) on degraded material, psi.
In the case of wall thinning, the capacity may be defined as
cr: (::;cr, )' +( ::, crs, )' +(!,cr.,)'

The pr obability of failure Pr is obtained from the relationship
where 8r = flow stress (e.g., 8r= (8,. + Su)/2 or l.lSy), psi
t:.t = wall thinning, in Pr 1 - (0.5 -t a rea under normal curve from O to J3).
t = nominal wall, in
The applied stress is the hoop stress; for example, in a cylindrical body 6.19 General Metal Loss Example
PD
A 40-in diameter x 0.875-in wall vessel has the following parameters:
Dem=-
2t Design pressure =540 psi (mean), 81 psi (std. deviation)
where P = design pressure, psi Flow stress = 57.8 ksi (meanl, 14.4 ksi (std. deviation)
D = outside diameter, in Wall t hinning = 0.15 in (mean), 0.15 in (std. deviation)
t = nominal wall, in
An inspection revealed a wall loss :lt = 0.10 in. Substituting, we obtain
The limit state function is therefore
1-l-g = 38.81 ksi
g=S ( 1 -~flt)_PD 15.68 ksi
f 3 t 2t
J3g 2.89
The parameters 8,, flt, and P are now probabilistic variables; in the
3
simplest case they can be viewed as normal distributions, each with a Pr = 6.68 X 10
mean f.l anda standard deviation cr. We define a reliability index J3, as
the ratio ofthe mean ofthe limit state function divided by its standard The probability of failure of the vessel is therefore estimated to be
deviation 6.68 in 1000, or 0.67 percent.
230 Chapter Six lnspectlon 231
6.20 Limit State Function for a Crack Design pressure = 540 psi (mean), 81 psi (std. deviation)
The limit state function and reliability index for a crack are calculated Residual stress= 24.5 ksi (mean), 24.5 ksi (std. deviation)
on the basis of fracture mechanics, presented in Chap. 9. Following the Fracture toughness = 80 ksi-fui (mean), 24 ksi-fui (std. deviation)
same approach as wall thinning but applied to a crack, the stress is
replaced by the stress intensity and the flow stress is replaced by the Geometry factor = 1.1 (mean), 0.5 (std. deviation)
fracture toughness. These parameters are explained in Chap.9. The Wall thinning = 0.15 in (mean), 0.15 in (std. deviation)
limit state function becomes Crack size a = 0.030 in (mean), 0.030 in (std. deviation)
g=KlC-(sr+~~)&Y The mean and standard deviation of the limit state function g are
g = 75.83 ksi--fui cr = 26.21 ksi.,,ID

-.!!L=1 The reliability index (3, for the crack is
dKrc
13 = 2.89
dg = .fiWy
dSr and the probability of failure is
Pr = 1.9 X 10-3
dg=D&Y
dP 2t For combined failure modes, if a series of degradation mechanisms
dg
da-
-(s + PD)Y
r
G_
2t 2 f~
is possible, each with its probability of failure Pli• then the total prob-
ability of failure is obtained as
If wall thinning and cracking are both possible, with a probability of

failure Pr1 and Pf2, respectively, the total probahility of failure is
obtained as
Pr = 1 - (1 - Pfl) (1 - Prz).
p= )lg
In the above examples, Pfl = 6.68 X 10-3 , and Pf2 = 1.9 x 10-3 ; then
Og
Pr =1- (1 - 6.68 X 10-3 )(1 - 1.9 X 10-3 ) = 8.57 X 10- 3
where Krc = fracture toughness, ksi..Jin
Sr = residual stress, ksi 6.22 Likelihood Reduction Options
P = design pressure, ksi
D = outside diameter, in The limit state function is useful in choosing among options to reduce
t = wall thickness, in the likelihood offai1ure. For example, ifwe want to reduce the likelihood
a = crack size, in offailure, calculated in Sec. 6.21, Pr = 8.57 x 10-3 , we could decide to
Y = geometry factor (Chap. 9) reduce the design pressure. A reduction of mean design pressure by 10
percent and a reduction of standard deviation Oack of accuracy) on the
pressure by 10 percent lead to
6.21 Crack Example
A 40-in diameter X 0.875-in wall vessel has the following parameters. P reduced = 90% P (Jreduced = 90% a
Repeating the calculation for Pfl and Pf2 we obtain the reduced prob- TABLE 6.7 Estlmate of likellhood L(nM)
abilities Low-reliability Moderate-reliability High-reliability
n prediction prediction prediction
P{l,red = 5.15 X 10- 3
Pf2. rt"<l = 1.76 X 10 3
Pr. red = 6.90 X 10- 3
1X 0.50 0.70 0.80

The probability of failure was 8.57 X 10 3 for a design pressure of 2X 0.30 0.2{) 0.15
540 psi; and now, with a 10 percent reduction in pressure and better
4X 0.20 0.10 0.05
accuracy, the failure probability is reduced to 6.90 X 10 3 , a reduction
of nearly 20 percent compared to the initial probability of failure of
8.57 X 10-3.
lf, instead of reducing the mean pressure by 10 percent, we obtain Given a method of predicting the wallloss M, the total probability of
actual material strength, and reduce the uncertainty (the standard failure is obtained by combining the individual probability of failures
deviation) on the flow stress <Js, to 40 percent ofthe initial14.4 ksi, the Pr multiplied by their likelihood L, in the form
probability of failure reduces significantly, down to
Pr = total failure probability
Pr - 1.91 X 10 3
= L(M) X Pr(M) + L(2M) X Pr{2tlt) + L(4M) X Pr(4M)
This is a reduction of 78 percent, a significant reduction, compared For example, ifthe calculation ofprobability offailure were conducted
to the initial probability of failure of 8.57 x 10-3 . as described in the preceding sections for a wallloss M, 2~t. and 4M,
and these were calculated to be
6.23 Correction for Reliability Pr(M) = 9 X 10'3 Pr (2M) = 20 X 10'3 Pr (3~t) = 200 X 10-3
The likelihood of failure of a vessel is calculated, as described above,
for a metalloss M. The calculation is then repeated for twice the metal and if the initial prediction of wallloss were based on prior inspections,
loss (2~t), and four times the metalloss (4~t). We call the correspond- which is considered a high-reliability method, then the total probability
ing probabilities Pr (M), Pr (2M), and Pr (4M). of failure is
Given an initial estimate of metal loss M, the likelihood that the Pr = 0.8 X 9 X 10'3 + 0.15 X 20 X 10-3 + 0.05 X 200 X 10·3 = 20 X 10'3
actual, measured metal loss is indeed M depends on the reliability of
the initial estimate.
6.24 Fully Quantitative Consequence
• A low-reliability initial estimate of wall loss is one that is based on
generic published data The five steps of a fully quantitative RBI consequence analysis are
illustrated in Fig. 6.5.
• A moderate-reliability initial estimate of wall loss is one that is
based on laboratory simulation or limited in situ coupons. Step 1. The scope could be a full system, a subsystem, or a single com-
• A high-reliability initial estimate of wallloss is one that is based on ponent such as the pressure vessel in Fig. 6.6, which shows, as flags,
extensive field data. the boundaries of the component or subsystem being evaluated. The
inventory ofthe vessel is 5000 lb ofhexane (CsH 14 ). The physical prop-
API 5814 provides estimates of the likelihood L(nM) of confirming an erties of hexane are:
initial estimate of wall thinning nM, as presented in Table 6.7. For
example, the likelihood that the measured wallloss is indeed 1 X M, Boiling point (°CI 69 (156°F)
when the initial prediction was 1 X M, is 50 percent i f the initial esti- Molecular weight 86
mate was of low reliability, and 80 percent if the initial estimate was Density relative to air 2.0
ofhigh reliability. The likelihood that the measured wallloss is 4 X ~t, Upper flammability limit (vol. 9é) 7.5
when the initial prediction was only 1 X M, is 20 percent if the initial
Lower flammability limit (vol.9é) 1.2
estimate was oflow reliability, and 5 percent ifthe initial estimate was
of high reliability. Heat of combustion fJ/kg) 4.5 X 107
1
System Scope
P&ID
Process Physical
Properties
~
2
Plant Map
Worker & Public
Density
~
Figure 6.6 Scope boundary of
vessel between flags.
3
Failure Modes
Event Tree
----- -- ... UVCE
~ : PB
4
Spill & Cloud
Radius
Risk Profile
O,' - - TF .- -- ---. '-,
, --.- - __ _-~F •• ":, , _BLEVE
v -... ' ·...

i
5
Quantified Risk
Figure 6.5
con~equence
Fully quantitative
analysis.
.
\_ o ' '-, ;.:,. B
0 .-0 -pF. .
r ---- \____ --------- .-

~·
-- ·• ' ; RD '.
•
- L - ------!
Ratio ofspecific hcats (k) 1.063

Liquid density at boiling point Ckg/m3 ) 615 R
5 Figure 6.7 Plant plot plan around ve ssel V.
Hcat of vaporization at boiling point CJ/kg) 3.4 X 10
Liquid heat capacity cJ/kg °KJ 2.4 X 103
• Off-plant. Population density, significant outside assets
Step 2. The plant plot plan is marked to indicate • In- and off-plant. Environmentally sensitíve areas
For example, for the purpose of illustration, Fig. 6. 7 is a very simplified
• In-plant. Office space, control rooms, zones ofhigh personnel density,
representation of the vessel V under evaluation, nearby road (RD), stor-
roads age building (SB), tank farm (TF), plant boundary (PB), and river (R).
• ln-plant. Adjacent flammable or toxic tanks or systems, critical safety The meteorology indicates wind direction distribution of 10 percent in
and emergency response assets · each ofthe following directions: N, NE, E, SE, and S; 15 percent in each
of the following directions: SW and W; and 20 percent NW; for a total 1.0E-6 . - - - - - - - - - - - - . . ,
of 100 percent.
(ij
8.0E-7 r-
Step 3. Failure modes and event tree. Three failure modes are con-
~
sidered: ....
Q)
o.
• Afull rupture ofthe vessel wall, with fullloss ofcontents, but no loss ~
o '--
from automatically isolable interconnected systems ,;: 1.0E-7
• A full rupture of the average nozzle in liquid space ~
• A vapor leak from a full rupture of a nozzle in vapor space
c
Q)
'C
'õ
u
An event tree is developed to indicate, for each scenario, the quantity <(
ofliquid or vapor that escapes from the postulated ruptures or leaks, the
liquid spill radius, the vapor cloud radius, and direcbon. Then the event
1.0E-8 +-.--------,----+----'
tree will estimate a probability for immediate or delayed ignition, result-
o 5 10
ing in ELEVE (boiling liquid vapor explosion), UVCE (unconfined vapor Damages, million $
cloud explosion), FF (flash tire), PF (pool tire), and safe dispersion.
Figure 6.9 Annual risk of a postulated failure.
Step 4. Spill and cloud radius and risk protile. The radü ofinfluence are
indicated in Fig. 6.7 for BLEVE, UVCE, FF, and PF. The corresponding causing $1M in damages is 8.0 X 10·7 per year. In industry a likelihood
consequence is plotted against distance from the vessel. The conse- less than 10·6 per year is generally considered an incredible event.
quence can be expressed as a protile of fire radius, heat radius, cost of
damage, or injury and fatality.
The consequence profile is then multiplied by the yearly likelihood 6.25 Advantages of Applying RBI
of the initiating event (e.g., vessel rupture) to obtain a risk profile, as • It provides a logical and clear method for setting inspection priorities.
illustrated in Fig. 6.8. The end result may be expressed, for example,
as individual risk of fatality per year versus distance from the vessel, or • It provides an industry benchmark for low-medium-high classification
as societal risk frequency ofN or more fatalities per year versus number of inspection priorities.
N offatalities. For example, in Fig. 6.9, the likelihood of an accident • It eliminates unnecessary inspections, and it ensures that essential
ones are performed.
• It fosters an intelligent dialogue and consensus process among key
disciplines: (1) inspection, (2) design, (3) materiais and corrosion, (4)
operations, and (5) safety.
• It improves knowledge ofprocess and material condition, leading to
improvements in reliability, safety, future designs, upgrades, and
maintenance.
6.26 Cautions When Applying RBI
Distance • Many RBI quantitative techniques only consider failures from over-
pressure and wall thinning; they may ignore other demands (fatigue
cycling, overload, etc.), and cracking mechanisms.
Figure 6.8 Risk profile. • RBI, as presented here, applies to static (stationary) equipment.
• The quality of the fmal product is as good as the input. This truism is • Re-inspection intervals, on the order offive to ten years, depending on
particularly valid when using software-based RBI. The key unknown results of inspections and operations, and regulatory requirements.
in an RBI assessment is corrosion and damage rate, primarily because
of changes in process stream and abnonnalloads (overpressure tran- 6.28 Overview of lnspection Techniques for
sients, abnormal start-up causing temperature transients, sudden Tanks, Vessels, and Pipes
and excessive vibration, unstable or explosive process, and so on).
Several nondestructive inspection techniques are available to inspect
• RBI assessment may not flag fabrication and construction flaws.
tanks, vessels, and piping. The choice of the technique depends on the
• With software-based RBI, the user must be familiar with the basis material, the type of flaw, access to the surface, availability, and cost;
ofthe corrosion modules (erosion, high temperature, etc.). these techniques include:
• RBI does take resources to develop, implement, and maintain.
• Visual examination
• Magnetic particle testing
6.27 lntegrity Programs for Pipelines
• Liquid penetrant testing
For safe and reliable operations, the oil and gas transmission pipeline
industry implements a systematic integrity management program UMP) • Radiographic testing
that combines pcriodic inspections, fitness-for-service evaluations, and • Ultrasonic testing
standard prequalified repairs. The IMP includes regulation and over- • Eddy current testing
sight from the Department ofTransportation, Office ofPipeline Safety.
In broad terms, Integrity Management P rograms for pipelines include • Acoustic emission testing
the following parts: • Magnetic flux leakage
• The preparation of a formal IMP plan, with inspection, reporting, and • Thermography
repatr stratcgy and responsibilities. • Laser profiling
• The definition of inspection priorities, starting with high-consequence • Replication
areas (HCAs), which include environmentally sensitive areas (drinking • Direct assessment
water, ecologically sensitive), high-population areas (over 1000 people
per square mile), and navigable waterways (commercial navigation). The keys to accurate inspections are:
• Selection of inspection strategy: in-line inspections (pig inspections),
hydrotesting, or direct assessment. • Inspection technique consistent with expected damage
• Protocol for data gathering, including original materiais, design and • Expertise and qualifications of inspectorsla.-12
construction data (i f available), operations and maintenance records, • Independence of inspector
and previous inspcctions. • Cleanliness of component
• Selection of inspection technique: pig inspection or direct assess- • Access to the component
ment of coating and cathodic protection.
• Good quality, calibrated instruments
• Method for data analysis and fitness-for-service; for pipelines these
techniques are typically ASME B31G or RSTRENG®, described in
Chap.7. 6.29 Visual Examination (VT)
• Leak detection and emergency response. Visual examination (visual testing, VT) is the most common examination
• Reporting. technique. It can be direct or assisted for remote access, for example,
through mirrors, borescopes, and cameras. Pipe Fabrication Institute
• Corrective actions, including pressure reduction, excavation, repairs,
Standard ES-2713 defines visual examination as examination with the
and mitigation (avoid recurrence).
"unaided eye," other than the use of corrective lenses, within 24 in.
Examiners are classified in increasing order of qualification from VT-1
to VT-3.
6.30 Magnetic Particle Testing (MT)

A magnetic field is created on the surface of the part, for example, by
using a yoke (Fig. 6.10), anda powder or solution ofmagnetic particles
is dispersed on the surface. The magnetic particles orient themselves
along the magnetic lines, and surface discontinuities become visible as
the magnetic lines appear disturbed.14 Wet fluorescent magnetic par-
ticles are particularly well suited for the examination of pipe, vessels,
and tank welds. The applicable standards are ASME B&PV Section V
Article 7, and ASTM standards.l5-1S
Advantages MT
• Detects surface and slightly subsurfacc flaws.
• Flaws do not have to be open to the surface.
• Portable.
Limitations MT
• Applies only to ferromagnetic materiais.
Figure 6.11 Liquid penetrant
• Surface technique cannot determine depth of flaw. reveals crack.l9
• Discontinuities detected if perpendicular to magnetic field.
• Surface must be sufficiently smooth to permit particle movement.
• Part may have to be demagnetized.
• No permanent record. 6.31 Liquid PenetrantTesting (PT)
• Inspection to be followed by vapor degreasing or chemical cleaning. A visible or fluorescent penetrant is applied to the surface for a few min-
utes (dwell time), during which time the penetrant seeps into surface-
connected flaws. The excess penetrant that did not penetrate the flaws
is then wiped away. Finally, a contrasting spray or powder developer is
applied to draw the penetrant back to the surface by capillary action.
This will outline the flaw shape (Fig. 6.11). The applicable standards are
ASME B&PV Section V Article 6, andASTM standards. 16· 2(}..23
Advantages PT
• Can be used on uneven surfaces.
• Portable.
Limitations PT
• Flaw must be open to the surface.
• Surface technique cannot determine depth of flaw.
• Affected by surface cleanliness, roughness.
• No permanent record.
Figure 6.10 Magnetic parti ele inspection of gouged surface.l4 • Inspection to be followed by vapor degreasing or chemical cleaning.
6.32 Radiographic Testing (RT)

A radiograph is a permanent image created by X-rays or gamma rays
passing through material. After a certain exposure time, a radiograph is
captured on radiographic film (classic radiography). The applicable stan-
dards for radiographic testing (RT) are ASME B&PV Section V Article 2,
and ASTM standards.l6,24--3.
The setup for RT includes t he selection of the X-r ay or gamma ray
source, the film, the filter to absorb the softer rays, the penetrameter
(image quality indicator), the distance source-object, the personnel exclu-
sion zone for radiation protection, and the exposure time which depends
on all the above factors, but is typically on the order of minutes for
radiography of tank, vessel, or pipe walls.
An evolution of conventional radiography is digital radiography.
Digital radiography relies on imaging plates and presents severa! advan-
tages over conventional radiography (Figs. 6.12 and 6.13). In particular,
digital radiography has a wide range of sensitivity, on the order of six Figure 6.13 Measurcment of wall thickness with digital radiography.
conventional films, which permits the examination of a large range of
thicknesses with a single plate. This, together with a shorter exposure
Advantages RT
time, no chemical processing time of the film, computer-based storage,
retrieval, and transmittal, makes it an attractive altemative to classic • Detects surface and volumetric flaws.
radiography. • Covers a relatively large area.
In real-time radiography, the radiographic image can be viewed • Provides a permanent record (film or digital).
ínstant1y on a screen as the object or the source is moved to inspect dif- • Recognized by construction codes.
ferent areas. • Detects narrow, cracklike flaws.
The qua1ity levei ofradiography is measured by penetrameters, such Limitations RT
as the "'2-2T" penetrameter that has a thickness ofT 2 percent ofthe • Requires a personnel exclusíon zone.
wall thickness, and a hole dia meter of 2T. • Requires experienced, certified operators.
• Requires an X-ray or gamma-ray (radioactive) source.
• Gamma rays have limited life.
• lt is difficult to decipher radiographies of complex shapes.
• Detects length of cracklike flaws, but may not characterize their
depth.
6.33 Ultrasonic Testing (UT)

A transducer in contact with the surface, through a couplant, emits
ultrasonic waves through the metal thick ness. The waves, either nor-
mal to the surface (straight beam) or at an angle (shear wave), reflect
off discontinuities or the opposite face (the pulse echo), indicating the
presence ofa flaw or the location ofthe opposite wall, and therefore the
wall thickness (Figs. 6.14 and 6.15). Alternatively, the pulse-emitting
transducer and the receiving t ransducer can be different (time offlight
Figure 6.12 Computer screen view of digital radiography image. diffraction, Fig. 6.16). The data are presented as scans:
Figure 6.14 A-scan of wall thickness between peaks.
Figure 6.16 TOFD ofjoint weld at

angle.38
• Takes time to measure a grid of many points, a wheel mounted

linear scanner may be used, Fig. 6.15.
• Multiple flaws may hide one another.
On-stream (in-service) UT inspection of high-temperature vessels
and piping represents several practical challenges:
Figure 6.15 Scanning transducer on wheels. • Regular transducers cannot sustain high temperatures. High-temper-
ature transducers contain a temperature-resistant stand-off material
between the crystals and the hot surface.
• A-scan is the raw signal on the instrument.
• Regular couplants evaporate at high temperature. More viscous,
• B-scan is the thickness profile. high-temperature couplants have been developed.
• C-scan is a plan view contour of the thickness of a whole region. • At high temperature, the signal tends to quickly disappear Oittle dwell
The applicable standards are ASME B&PV Section V Articles 4 (in- time). A freeze function on the instrument helps solve the problem.
service examination) and Article 5 (examination of materiais and fab- • In hot metal, the velocity of sound decreases and signal attenuation
rication), and ASTM standards.l6,32-37 increases as a result of the higher molecular movement. This can be
resolved either by placing the calibration block on the component
Advantages UT surface and allowing it to heat to metal temperature or, preferably,
• Only one side needs to be accessed. correcting the readings through charts or formulas such as found in
• Can be used on complex shapes. Ref. 39.
Limitations UT • As the transducer's stand-offmaterial heats it expands, which affects
• Requires a smooth surface. the zero offset reading, causing a higher than actual reading of the
• Requires a couplant. component wall thickness. This can also be corrected after recording
• Leaves no permanent record. the reading.
Figure 6.18 Guided wave UT.40.41
Figure 6.17 UT grid.l9
6.34 Long-Range Guided Wave

Ultrasonic lnspection
A collar with transducers is wrapped around the pipe as shown in Fig.
6.18. The transducers send ultrasonic waves longitudinally along the
pipe, and can detect inner and outer wall loss. The signal from a single
collar reaches 100 ft or more along the pipe.
6.35 Eddy Current Testing

Eddy currents are generated by a probe into the wall of a specimen.
The presence of a flaw in the wall ora change in wall thickness will be
detected by a disturbance ofthe current. The applicable standards are Figure 6.19 Pulsed eddy current testing through-insulation.s
ASME B&PV Section V Article 8 and ASTM standards.I6,42
Pulsed eddy currents are used to measure wall thinning under insula-
tion (Fig. 6. 19).4:! Changes in wall thickness can be detected on pipe wall to flow in any nearby wall of a metallic component, plate or pipe. The
thickness from 0.3 in to 1.5 in, with accuracy on the order of a few mils. magnetic and electric signals are affected by the distance between
This ability to measure wall thickness while the line is in service and the coil and the metal, and the thickness of the metal. Wall thinning and
without removing insulation can lead to significant cost savings. pitting can then be detected by the loss and distortion of the magnetic
and electric signals. This is the principie at the basis ofthe magnetic flux
leakage (MFL) tools common in pipeline pigs and also used to inspect
6.36 Magnetic Flux Leakage tank walls and bottom plates. Other applications based on magnetic field
An alternating electric cunent circulates in a coil, creating a magnetic and eddy currents have been developed to characterize and map surface
field. In turn, the magnetic field causes reaction currents, eddy currents, pitting.44
248 Cha pte r Six lnspection 249
6.37 Acoustic Emission Testing (AE) and filled with liquid (couplant). The tool can be lowered by gravity
and pulled back up, or it can be equipped with a remotely controlled
Under increasing stress, the tip of a crack plastically deforms, and if motor to inspect horizontal runs.51
the stress continues to increase, the crack will grow. The plastic defor-
mation as well as the crack growth can be detected as acoustic signals
emanating from the cr ack. Th e principie of acou stic emission (AE) is to 6.39 Leak Detection Systems
detect these acoustic signals. For that, a n array of piezoelectric tra ns- Leak detection in liquid or gas systems can be achieved through an anay
ducers is placed on the component wall, and the component is pressur- of inline or clamp-on meters that record flow rate and pressure. The
ized. The transducers detect stress waves from cracks, and pinpoint t he flowmeters are complemented by software that continuously processes,
location of the flaw by triangulation. As in ultrasonic testing, the trans- monitors, and compares flow parameters at various points along the
ducers convert the ultrasonic emission into an electric signal; the signal network, and provides alarms when sensing a mass imbalance that
is sent to a filter, an amplifier, and an analyzer. The source of acoustic would indicate a leak. Some systems can also measure density and
signals is located by triangulation, and more detailed inspection of the tempera ture and therefore can identify interfaces between different
source can then be performed by UT. AE has also been used to detect batches of liquids.52
cracking during or subsequent to weldin g.
The applicable standards are ASME B&PV Section V Articles 11, 12,
and 13, and ASTM standards.l6,45-49 6.40 Direct Assessment
Some pipelines, particularly gas pipelines, are not pigable for reasons
AE advantages such as multiple pipe size, bends, and valve opening restrictions. To
• Only one side needs to be accessed. assess the integnty of these pipelines, ASME has published ASME
• Sensors can cover a very wide region in one test. B31.8S, Managing System Integrity o{Gas Pipelines, 53 which provides
• May inspect different components in a single test. altematives to pigging. These altemate integrity assessment tech-
• May be performed in-service, or during cool-down or shutdown. niques include:
• Detects growing cracks, not stable construction flaws.
• Excavation to detect coating defects, surface damage (VT, pit gauge,
AE limitations PT, MT, portable MFL), wall loss (UT), and cracks (angle beam UT,
• Crack may fail through-wall during pressurization test. guided wave)
• Requires specialized services and expertise.
• Excavation for soil char acterization, resistivity, and conosivity surveys
• Coating surveys
6.38 Pig lnspections of Pipelines
• Cathodic protection surveys
A pig is a tool that moves through a pipeline for the purpose of cleaning
and drying Cutility pigs) or inspection and dimensioning (smart pigs).50
References
• Utility p1gs include swab pigs, mandrel pigs, foam pigs, gauging 1. Geitner, F. K, Setting inspection frequencies, PipeLine and Gas Technology,
pigs, plug pigs, dewatering pigs, and batching pigs. November/December 2004.
2. 49 CFR Transportation, Parl 192, Transportat10n of Natural Gas mui Other. Gas by
• Smart pigs include ovality tools, gyroscopic pigs, magnetic flux leak- Pipeline: Mininwm Federal Sa(ety rgas pipelines, ASME B31.8l. Part 193, Ltque(ied
age pigs (MFL), and ultrasonic pigs. Natural Gas Facilities: Federal Safety Standards. Part 194, Response Plans for
Onshot·e Oil Pipelines. Part 195, Transportation of Hazardous Liquids Pipelines,
In large pipelines, pigs are product-propelled by the flow of gas or Code of Federal Regulations. Washington, DC.
3. API RP 580, Risk·Based Inspection, American Petroleum Inst.itute, Washington, DC.
liquid, with the pig launched at one point and collected at a trap down- 4. API Publ. 581, Risk·Ba!!ed Inspection Base Resource Document, American Petroleum
stream. But when there is only one access point, cable-operated (teth- Institute, Washington, DC.
ered) pigs have been successfully used. These pigs can be articulated 5. Leonard, R.. Anderson. S .. Risk-based Inspection a Pilot Project Overview, in ASME
to clear tight bends, and they rely on ultrasonic standofftools that per- PVP Conference, 1998, San Diego. . .
6. ASME, Post·Con.'<truction Code, Inspection Plannmg, under development, Amencan
mit the tool to pass changes in diameter. The pipe has to first be cleaned Society ofMechanical Engineers, New York.
7. NBIC ANSI/NB-23, National Board oflnspection Code, The National Board ofBoiler 34. ASTM E 213, Standard Practice for Ultrasonic Examination of Metal Pipe and
and Pressure Vessel Inspectors, Columbus, OH. Tubing, American Society for Testing and Materiais, West Conshohocken, PA.
8. Antaki, G. A., Monahon, T. M., Cansler, R. W., Risk-based Inspection ofSteam Systems, 35. ASTM E 428, Standard Practice for Fabrication and Contrai of Steel Reference
in ASME PVP Conference, 2005, Denver. Blocks Used in Ultrasonic Examination, American Society for Testing and
9. Kirsner, W., Surviv-ing a steam rupture in an enclosed space, heating, piping and air Materiais, West Conshohocken, PA.
conditioning, HPAC Engineering, July, 1999. 36. ASTM E 500, Standard Terminology Relatin.g to Ultrasonic Examination, American
10. ASNT, The Nondestructiue Testing Handbook on Radiography and Radiation Testing, Society for Testing and Materiais, West Conshohocken, PA.
American Society ofNondestructive Testing, Columbus, OH. 37. ASTM E 797, Standard Practice for Measuring Thíckness by Manual Ultrasonic
11. ASNT 2055, RecomTTUinded Practice No. SNT-TC-lA, American Society ofNondestructive Pulse-Echo Con.tact Method, American Society for Testing and Materiais, West
Testing, Columbus, OH. Conshohocken, PA.
12. ASNT 2505, Standard for Qualifi.cation and Certifi.cation of Nondestructiue Testing 38. RTD Quality Services, Edmonton, Alberta, Canada.
Personnel, also known as CP-189, American Society of Nondestructive Testing, 39. Nisbet, R. T., Ultrasonic Thickness Measurements at High Temperatures, The NDT
Columbus, OH. Technician, October, 2004.
13. PFI Standard ES-27, Visual Examination, The Purpose, Meaning and Limitation of 40. Petrochem Inspection Services, Houston.
the Term, Pipe Fabrication Institute, Springdaie, PA. 41. Lebsack, S., Noninvasive lnspection Method for Unpiggable Pipeline Sections,
14. Kiefner & Associates, Worthington, OH. Pipeline & Gas Journal, June 2002, pipelinegasjoumaionline.com
15. Huber, O. J., Fundamentals of Nondestructiue Testing, Metais Engineering ilnstitute, 42. ASTM E 243, Practice for Electromagnetic (Eddy-Current) Testing of Seamless
Metais Park, Ohio, 1984. Copper and Copper-Alloy Tubes, American Society for Testing and Materiais, West
16. ASME V, Nondestructiue Examination, American Society of Mechanical Engineers, Conshohocken,PPL
New York. 43. Cohn, M. J ., de Raad, J. A., Nonintrusiue Inspection for Flow-Accelerated Corrosion
17. ASTM E 269, Standard Defínitions of Terms Relating to Magnetic Particle Detection, PVP-Vol. 359, ASME 1997, American Society of Mechanicai Engineers,
Examination, American Society for Testing and Materiais, West Conshohocken, PA. NewYork.
18. ASTM E 709, Practice for Magnetic Particle Examination, American Society for 44. Couch, A., et ai., New Method Uses Conformable Array to Map Externai Pi.peline
Testing and Materiais, West Conshohocken, PA. Corrosion, Oil & Gas Journal, Nov. 1, 2004.
19. Thieisch Engineering, Cranston, RI. 45. ASTM E 569, Standard Practice for Acoustic Emission Monitoring of Structures
20. ASTM E 165, Liquid Penetrant Inspection Method, American Society for Testing and During Controlled Stimulation, American Society for Testing and Materiais, West
Materiais, West Conshohocken, PA. Conshohocken, PA.
21. ASTM E 260, Standard Practice for Packed Column Gas Chromatography, American 46. ASTM E 749, Standard Practice for Acoustic Emission Monitoring During Welding,
Society for Testing and Materiais, West Conshohocken, PA. American Society for Testing and Materials, West Conshohocken, PA.
22. ASTM E 433, Standard Reference Photographs for Liquid Penetrant Inspection, 47. ASTM E 751, Standard Practice for Acoustic Emission Monitoring During
American Society for Testing and Materiais, West Conshohocken, PA. Resistance Spot Welding, American Society for Testing and Mat.erials, West
23. ASTM E 1417, Standard Practice for Liquid Penetrant Examination, American Conshohocken,PA.
Society for Testing and Materiais, West Conshohocken, PA. 48. ASTM E 1067, Standard Pmctice for Acoustic Emission Examination of Fiberglass
24. ASTM E 94, Standard Guide for Radiographi.c Testing, American Society for Testing Reinforced Plastic Resin, American Society for Testing and Materiais, West
and Materiais, West Conshohocken, PA. Conshohocken, PA.
25. ASTM E 142, Standard Method for Controlling Quality of Radiographic Testing, 49. ASTM E 1118, Standard Practice for Acoustic Emission Examination of Reinforced
American Society for Testing and Materiais, West Conshohocken, PA. Thermosetting Resin Pipe, American Society for Testing and Materiais, West
26. ASTM E 242, Standard Reference Radiographs for Appearances of Radiographic Conshohocken,PA.
Images as Certain Parameters Are Changed, American Society for Testing and 50. NACE, Standard RP0102, In-Line Inspection of Pipelines, NACE International,
Materiais, West Conshohocken, PA. Houston.
27. ASTM E 747, Standard Practice for Design, Manufacture, and Material Grouping 51. Van Agthoven, R., de Raad, J. A, Fieid Experience Shows How to Inspect Odd,
Classifícation of Wire lmage Quality Indicators (IQ[) Used for Radiology, American Noninspectabie Platform Risers, Oil & Gas Joumal, February 9, 2004.
Society for Testing and Materiais, West Conshohocken, PA. 52. Controlotron Corporation, Hauppauge, NY.
28. ASTM E 999, Standard Guide for Controlling the Quality ofIndustrial Radiographic 53. ASME B31.8S, Managing System Integrity of Gas Pipelines, American Society of
Film Processing, A.merican Society for Testing and Mate,;als, West Conshohocken, Mechanical Engineers, New York.
PA.
29. ASTM E 1025, Standard Practice for Design, Manufacture, and Material Grouping
Classifi.cation of Hole-Type lmage Quality lndicators ([Ql) Used for Radiology,
American Society for Testing and Materiais, West Conshohocken, PA.
30. ASTM E 1030, Standard Test Method for Ra.diographic Examination of Metallic
Castings, American Society for Testing and Materiais, West Conshohocken, PA.
31. ASTM E 1079, Standard Practice for Calibration of Transmission Densitometers,
American Society for Testing and Materiais, West Conshohocken, PA.
32. ASTM E 114, Standard Practice for Ultrasonic Pulse-Echo Straight-Beam
Examination by the Contact Method, American Society for Testing and Materiais,
West Conshohocken, PA.
33. ASTM E 164, Standard Practice for Ultrasonic Contact Examination of Weldments,
American Society for Testing and Materials, West Conshohocken, PA.
Chapter
7
Thinning
7.1 Three Categories ofWaiiThinning

For the purpose offitness-for-service, wall thinning can be divided into
three categories:
• General metalloss (GML)
• Local metalloss (also referred to as local thin area LTA)
• Pitting
The fitness-for-service evaluations for metalloss consist oftwo checks:
• Rupture prevention. Check that the corroded component has sufficient
strength to resist applied loads (pressure, weight, temperature, and
so on).
• Leak prevention. Check that the remaining ligament is sufficiently
thick to prevent pin-hole leaks.
The methods for ruptw·e prevention of general and local metal loss
are closely related. They are based on the concept of metal reinforce-
ment in ductile material: the weak thinner metal area (WM in Fig. 7.1)
is reinforced by the surrounding sound metal (SM in Fig. 7.1) provided
the thin region WM is not too large.
Rupture prevention for pitted components takes a different
approach. A pitted metal is evaluated on the basis of ASME VIII, Div.1,
Appendix AA, which provides rules for tube sheets (strength of liga-
ment in perforated plates t. The pits are considered to be perforations
and the assessment evaluates the strength of the remaining wall
between pits. In addition, in all three cases (GML, LTA, and pitting)
253
254 Chapter Seven Thinning 255
~ t t • -rrtt t t n---r-r t p t t ~
SM WM SM
Figure 7.1 Local wall thickness reinforced by adjacent metal.
the remaining wall thickness is kept above a minimum threshold to

prevent the formation of pin-hole leaks.
I
7.2 Leak or Break I
I
f--i------~--:-~----~----~
The fitness-for-service assessment will differentiate between criteria
to prevent a leak and criteria to prevent a break:
• Leakage will occur when the remaining metal wall is on average suffi-
cient to sustain the stresses due to operating loads, but is insufficient
tnom 1 14 ~ I tmin
to prevent pin-hole leaks. Referring to Fig. 7.1, this means that WM I S I
is small.
Figure 7.2 Wall thinning parameters.
• Rupture from wall thinning will take place when wall thinning
occurs uniformly, to the point that a whole region can no longer sus-
tain the operating stress. Referring to Fig. 7.1, this means that WM For example, the minimum required wall thickness of a boiler head
is large. is calculated following the rules of ASME Section I, to be 0.9 in. To this
thickness the designer adds 0.3 in corrosion allowance, and buys a
In summary, and this is quantified later, leakage from wall thinning head that is 1.25-in thick, slightly more than the 0.9 + 0.3 = 1.20 in
occurs if the metal is thick with localized thin spots. A break will occur needed. The nominal wall thickness of the head is tnom = 1.25 in, and
if the metal is down to a thin membrane over a large area. This logic the minimum required wall thickness is trnin = 0.9 in.
assumes two things: Next, we delineate the length of the zone where the metalloss has
encroached on trnin; we label this length S.
• There is no violent overload that could rupture even the original The length S is to be measured longitudinally to assess the integrity
uncorroded wall ( Chap. 11). of the component under pressure, and it is to be measured along the cir-
• The material behaves in a ductile manner ( Chap. 2). cumference to assess the integrity of the component under bending and
axialloads. In other words, S has to be measured in the direction per-
pendicular to the stress caused by the applied load.
7.3 When ls Corrosion Considered General?
In Fig. 7.3 the dominant load is internai pressure, the maximum
Corrosion is general if the extent of wall thinning below the code stress dueto pressure is circumferential (hoop) <Jc, and the length of
required minimum thickness tmin. labeled S and shown in Fig. 7.2, is corrosion ofinterest is therefore longitudinal. In Fig. 7.4 the dominant
larger than a certain limit length L calculated later in this chapter. load is bending, the stress due to bending <JL is longitudinal (axial), and
The method of API 579 is followed to defineS and L. 1 We first calcu- the length of corrosion of interest is therefore circumferential. 'Ib trace
late the minimum wall thickness that would be required by the design the line through the grid, imagine that the corroded region is a canyon,
and construction code; this m.inimum required thickness is tmin• and is and the line is a river at the bottom of the canyon; the riverbed is the
indicated in Fig. 7.2. The minimum required wall thick.ness tmm is of line through the minimum wall thickness points. This riverbed line is
course lower than the original nominal thickness tnam· called the criticai thickness profile (CTP).
tmin = minimum wall required by design code, in

RSFa = remaining strength factor, allowable, explained in Sec. 7.22
• lf S >L, as is the case in Fig. 7.2, then the metalloss is considered
to be general, and the assessment ofthe corrosion zone continues fol-
I I I I I I
lowing the general metalloss (GML) rules.
I I I I I I
I I I I I I • If S <L, the corroded zone can be assessed as a local thin area (LTA).
I
~
t I I
7.4 Principies of Evaluation for GML
FigYre 7.3 CTP for circumferential stress crç. The evaluation of general metalloss for tanks, piping, and pressure
vessels consists of averaging the measured remaining wall thickness
over the length of thickness averaging L, and then applying three
acceptance criteria to the averaged minimum tam:
--- - -
• The averaged minimum thickness in the corroded area tam minus the
future corrosion allowance, must be larger than tmin:
i
tam - FCA ~ tmin.
-- -----
---------- ----- -- ------ This criterion is intended to verify that the corroded component
--- ------- can sustain the internai pressure. This criterion protects against
rupture of a ductile material, where rupture is preceded by outward
bulging ofthe thinned area under large hoop stress, as illustrated in
Fig. 7.5.
Figure 7.4 CTP for longitudinal stress CfL.
• The minimum measured thickness tmm, minus the future corrosion
allowance, must not be thinner than 0.10 in:
Next, we calculate a length, labeled the "length ofthickness averaging"
L, which is illustrated in several examples that follow, and is given by tmm- FCA ~ 0.10 in
This criterion is intended to prevent pinhole leaks.

L=QJDt
Q = 1.123 ( 1-Rt
1-Rt /RSF,
)z -1
R= tmm -FCA
t t mm
.
where D = diameter, in
Q = parameter
Rt = remaining thickness ratio
FCA = future corrosion allowance, in
lmm = minimum measured remaining wall, in Figure 7.5 Overpressure failure of ductile materiaJ.2
258 Chapter Seven Thinnlng 259
• The minimum measured thickness tmm, minus the future corrosion

allowance, must not be thinner than halfthe original wall thickness:
t -FCA > tnom

mm - 2
This criterion is intended to flag out unusually large wall loss. An

alternative would be to limit tmm to 20 percent lnom as in ASME B31G
pipeline integrity criteria. 3 API 579 uses tmin rather than tnom·
7.5 Limitations
The assessment of wall thinning, as described in this chapter, is only
valid under the following conditions.
• The equipment is not operating in the creep regime, as defined in

Chap. 10.
Figure 7.6 Buckling due to wall thinning.
• There are no gouges (knife-edge cuts), grooves (sharply localized cor-
rosion), or cracks in or around the thinned area.
• The thmned area is at least a distance ...fl5i away from a "structural - J "J.. :s l.f l:> ~ -,. c· '-{
f
I~
discontinuity," where D is the component diameter and tis its thick-
A\
ness. Structural discontinuities are abrupt changes in geometry as
encountered at shell-to-flange or shell-to-support attachment. Ifthere
is a structural discontinuity. fitness-for-service of the thinned area
must be based on finite element analysis. which is a Levei 3 assess-
ment, as described in Chap. 1.
6'
c..
r: :
..
.
7.6 Buckling .D o
Another consequence of wall thinning may be buckling. It is a clear

concern for tall towers due to the weight bearing down on the thinned
area, but it is also a concern if thinning occurs at bearing points; in
Fig. 7.6, the tank buckled at the saddle support dueto wall thinning. Figure 7.7 Half-inch grid spacing for ultrasonic inspection.
Interestingly, in this case, the buckled tank did not leak.
For example, in a shell 48-in OD X 0.5-in wall, the grid spacing would
7.7 Ultrasonic Grid
be
A fitness-for-service assessment for wall thinning cannot be conducted
on the basis of a single point. Ideally, a continuous scan of the wall thick- GS - 2 X 0.5 = 1 in
ness should be obtained to delineate the area of wall loss. As a mini-
muro, straight beam ultrasonic thickness readings should be recorded 7.8 Storage Tank Example
on a grid, as indicated in Fig. 7.7, with spacing equal to
A storage tank has a height of 50 ft and a diameter of 50 ft. It was fab-
GS = 2tnom ricated of ASTM A 516 Grade 55 carbon-manganese steel plate. The
shell is 0.55-in thick at the bottom course, and the tank welds were
where GS =grid spacing. in, and tnom = nominal wall thickness. in. inspected by spot radiography. The tank was originally designed for 30
260 Chapter Seven Thlnnlng 261
n_,~
years of service, with a design corrosion rate of 5 mils/yr. The design fill
height is 45 ft, maximum, with a fluid that has a specific gravity of 1.4.
After ten years of service, a straight beam ultrasonic inspection of
the bottom course indicated significant corrosion under the insulation,
ali around the bottom of the tank (Fig. 7.8 '· The minimum remaining
-=1 L. • tnlll
wall thickness was measured to be 0.25 in. The worst, most corroded,
profile (referred to as criticai thickness profile or CTP) is shown in Figs.
7.9 and 7 .10. This corroded zone occurs 2ft from the tank bottom, which
is a "major structural discontinuity." 0.35
We need to know whether the tank can be kept in service, as-is, for
0.35
one more year.
We first calculate the minimum wall thickness required by the applic- 0.25
s
able construction code, in this case API 650. 4 The '"one-foot method" of 0.25
API 650 is followed to calculate the minimum required shell thickness
10 readings 0.25
@ 1'
0.25
0.25
2.6Df H -1)G CA
td = + 0.25
sd
0.30
t =_2._6D_.;_(H_-__:_1)
, s I
0.30
where tmm,li!MJ = minimum wall required by API 650

Figure 7.9 Corrosion profile. Figure 7.10 Criticai thickness profile.
td = design shell thickness, in

t1 = hydrostatic test shell thickness, in
D = nominal tank diameter, ft
H = design liquid levei relative to bottom of course, ft
G = design specific gravity of contents
CA = corrosion allowance, in
sd = allowable stress for the design conditions, psi
8 1 = allowable stress for the hydrostatic test condition, psi
API 650 defines the allowable stresses Sd and 8 1 as
Sd = min( ~S.v;~Su)
S, = min( ~S1 ;~S~)
Figure 7.8 General \;ew of corroded bottom tank course. where S .. = yield stress, psi, and S. ultimate strength, psi.
The strength properties and allowable stresses of ASTM A 516 Grade 7.1 O Tank Fitness-for-Service-Level 1
55 carbon-manganese steel plate, at 70°F ambient temperature, are S,.
= 30 ksi and Su = 55 ksi. Therefore Sd = 20 ksi and S, = 22.5 ksi. The Before we proceed with an API 579 Level1 fitness-for-service evaluation
specific gravity of the stored material is G = 1.4. The design fill height, for general metal loss (GML), we check the conditions of applicability
relative to the bottom of the tan.k is H = 45 ft. from Sec. 7.5. In particular, the corroded area should be at least ,I5i from
At the original design stage, the design corrosion allowance was set the base.
at 0.005 in (5 mils) per year for 30 years, CA = 30 x 0.005 in= 0.15 in.
Therefore, the design wall thickness of the bottom course was calculated JDi = J600 X 0.55 = 18 in
to be
Because the distance from the corroded zone to the tank bottom, 24 in,
_ 2.6 X 50 X (45 - 1) X 1.4 O _O . is larger than 18 in, the corroded zone is not considered near a struc-
td - 20,000 -r · 15 - ·55 10 tural discontinuity. What this means in practice is that the stresses in
the tan.k wall are governed by the hydrostatic pressure of the fluid; at
t = 2.6x50x(45-1) = 0.25 in this distance from the bottom there are no significant local bending or
' 22,500 peak stresses due to nearby changes in shape or boundary conditions.
The minimum wall thickness required by the construction code,
tmm.650 = max(0.55 in;0.35 in)= 0.55 in excluding corrosion allowance, is
We also verify that the design thickness of 0.55 in is not less than tmin,650 = 0.55 in - 0.15 in - 0.40 in
the API 650 minimum shell thickness, which for a 50-ft tank is 0.25 in.
The minimum measured wall thickness is tmm = 0.25 in, where tmm
= minimum measured remaining wall, in.
7.9 API 653 Tank Thickness The remaining thickness ratio R" the ratio of the worst projected
For comparison to API 650, the minimum design shell thickness of a corroded thickness over the minimum wall required by design code, isl
storage tank, according to API 653, is5
_ tmm - FCA
RI -
2.6DCH -1JG fmtn
tmin,653 = SE
where R, = remaining thickness ratio, and FCA = future corrosion
where S = maximum allowable stress, psi, and E = tank weld joint allowance, in.
efficiency.
The tan.k was originally designed assuming 5 milslyr corrosion rate.
The maximum allowable stress is Instead, the tank bottom actually corroded 0.25 in within ten years, a
corrosion rate of25 milslyr. We want to know ifit can remain in service
S = min(0.80 x Sy; 0.426 x S ..) = min(0.80 x 30; 0.426 x 55) = 23.4 ksi
for one more year, with a projected additional corrosion of 25 mils, or
The weld joint efficiency is 0.85 if, during construction, the tank FCA = 0.025 in. The remaining thickness ratio is ther efore
welds were inspected by spot radiography, and O. 70 if the welds were
not radiogra phed.5 In our example, R = 0.25- 0.025 = 0.56
I 0.40
t . = 2.6 X 50 X ( 45 -1) X 1.4 = O in
40
mm,S53 23, 400 X 0.85 . The parameter Q is given byl
l-~F , )
2 2
Therefore, the measured corrosion profile, with 0.25 in remaining of 0 56
the original 0.55-in thick bottom course, is below the 0.40-in minimum Q =1.123 ( - 1=1.123 ( l- · ) -1 =0.66
l - RI 1-0.56 0.90
corroded wall pennitted in API 653, unacceptable according to API 653.
The "length of thickness averaging" is the length of metal over which In our example, 0.25 in - 0.025 in is not :2: 0.5 X 0.55 in = 0.275 i n.
the wall thickness readings may be averaged, it is given by1 Therefore, the corroded tank does not meet Conditions 1 and 3 ofLevel
1 fitness-for-service.
L= QJDtmin,650 = 0.66.j(50 X 12)x 0.40 = 10 in
7.11 Resolution
Because the longitudinal extent ofwall thinning below tmm.650 = 0.40
in extends over a distance s of 12 in, which is larger than L, the corrosion One possible soiution would be to Iower the tank level to meet
condition is classified as a general metal loss. For this metalloss to be Condition 1, the strength condition, based not on the minimum mea-
acceptable, the three conditions of Sec. 7.4 must be met: sured wall tmm = 0.25 in but on the averaged minimum tam = 0.28 in.
Condition 1. The first condition is a remaining strength condition; Applying the minimum wall condition we obtain
it will check whether the corroded wall (current wall thickness minus
future corrosion) has sufficient strength to hold the internai pressure. 2.6DCH -l)G
To verify this condition, we calculate t h e corroded wall thickness, along tam = SE
the worst profile (the criticai thick.ness profile), averaged over the
length of thickness averaging L = 10 in. Referring to Fig. 7.10, within H= (t~- FCA)xS + = (0.28 in -0.025 in)x 20,000 + 1 = 29 ft
a 10-in distance of the worst reading of 0.25 in, the average wall is 1
2.6 X D X G 2.6 X 50 X 1.4
0.35 in + 0.35 in + 0.30 in+ 0.30 in+ 6 x 0.25 in . The content levei would have to be reduced to 29ft from a design
t = = 0 .28 ln
~ 10 height of 45 ft. This, however, is insufficient because Condition 3 is
still not met. The corrosion rate is simply too large; there is a possibil-
Then, we deduct the future corrosion allowance 0.28 in - 0.025 in ity that a pinhole leak or rupture may develop because Condition 3
0.255 in. Condition 1 can be written as t=> - FCA :2: taun.650· In our is not met.
exampie, this condition is not met because 0.255 in < 0.40 in. The choice is now to either repair lhe corroded lower tank course, or
Therefore. the first condition is not met. At this point, there are three proceed to a Levei 2 evaluation.
options:
• The fitness-for-service assessment proceeds to Levei 2 . 7.12 Tank Fitness-for-Service-Level 2

• The fill height is reduced. The Levei 2 assessment for fitness-for-service is very similar to
• The corroded bottom shell course is repaired. that ofLevell. The oniy difference 1s that Conditions 1 and 2 intro-
duce the remaining strength facto r allowable RSF" 1 described in
For the purpose of illustration, we check the second condition of the Sec.7.3.
Levei 1 assessment. CondltlOn 1, Leuel2.
Condition 2. The second condition will check that there is no poten-
tial for pinhole Ieaks. The potential for pinhole leaks is prevented if the t an•- FCA? RSFa X tmin.s5o
remaining wall minus the future corrosion allowance is at least 0.10 in
With an RSFa of 0 .9, this condition is still not met because 0.28 in-
(0.25 mm) thick. Condition 2 can therefore be written as tmm - FCA :2:
0.025 in < 0.90 X 0.40 in.
0.10 in. In our example this condition is met because 0.25 in- 0.025 in
Conditions 2 and 3 remain unchanged for Levei 2.
:2:0.10 in.
Condition 3. The third condition places a limit on corrosion rate; it lmm - FCA :2: 0.10 in
requtres that the remaining wall be at least half the nominal wall
t -FCA > tnnm
mm - 2
t -FCA > tnom
mm - 2 Level2 does not pass because Conditions 1 and 3 are still not met.
t . = PD = 130x22 =0.12 in
mm 2(SE + PY) 2(12, 000 X 1.0 + 130 X 0.4)
where P = internai design pressure; for the purpose of investigation

we use the operating pressure, psi
S = code allowable stress, from ASME B31.1 Appendix A, for
ASTM A 53 seamless carbon steel at 280°F
E = weld joint efficiency = 1.0 for a seamless pipe
y = 0.4 for ferritic steel below 900°F
The minimum measured wall thickness is lrnm = 0.024 in. The remain-
ing thickness ratio is
Rt = tmm -FCA = 0.024-0 =0.2

tmin 0.12
The parameter Q is
2 2
1-R 1-0.20 -1 =0.26
Q=1.123 ( 1 1
_R /~Fa ) -1=1.123 ( 1-0.20 I 0.90 )
Figure 7.11 Ruptured plant pipe.
The length of thickness averaging is
7.1 3 Power Plant Pipe Rupture
A 22-in OD X 0.39-in wall, carbon steel piping system, part of a plant L= Q~Dtmin = 0.26-./22 X 0.12 = 0.4 in
utility system, is not classified as essential and so is not part of the
plant's inspection program. The system contains water at 280°F, pressur- We therefore can average the readings over only 0.4 in, which means,
ized at 130 psi to remain liquid, and flowing at approximately 6 ftlsec. in practice, that we do not average the thickness. We have to use the
One day, after many years of service, the line suddenly and violently thinnest point for the fitness-for-service assessment. Note that in this
ruptured (Fig. 7.11). The rupture was wide, with the 280°F water flash- particular case, it was not necessary to calculate L because averaging
ing into steam as it escaped the pipe. Upon inspection, it was reported the lowest reading with surrounding metal thickness would not have
that the line had thinned down to 24 mils from its original 390-mils wall helped inasmuch as the metal was uniformly thinned down to 0.024 in
thickness, and this thinning had taken place over a wide area. over a wide area.
The condition ofthe pipe and its fitness-for-service are investigated in lgnoring, for now, the longitudinal stresses dueto bending, and look-
the following sections, assuming that the pipe was ASTM A 53, Type S ing only at the hoop stress dueto the operating pressure, the first condi-
(seamless), Grade B carbon steel, with the mechanical properties Sy = tion for a Level1 fitness-for-service condition is tam - FCA ~ tmin which
35,000 psi and Su = 60,000 psi, where Sr = minimum specified yield is not met because 0.024 in - O < 0.12 in.
stress, psi and Su = minimum specified ultimate stress, psi. The hoop stress in the uniformly thinned section is
7.14 Power Pipe Fitness-for-Service--Level1 PD = 130 x 22 = 59 580 si

2tmm 2 X 0.024 ' p
The minimum ASME B31.1 code required wall thickness to sustain a
pressure of 130 psi at 280°F, in an ASTM A 53 seamless carbon steel This hoop stress is practically equal to the ultimate strength of the mate-
pipe, is6 rial, and would explain the rupture by itself, even without consideration
of the supplementary longitudinal stresses caused by bending and The bending stress in the wall thinned region is quite high, at nearly 70
internai pressure. percent of the ultimate strength of the material.
To evaluate the fitness-for-service of components subject to longitu- In conclusion, the corroded region does not pass Levei 1, and would
dinal stresses (typically longitudinal pressure stresses plus applied not pass either Leveis 2 or 3 of a fitness-for-service evaluation. The
bending stresses), we must consider the criticai thickness profile perpen- internai pressure alone brings the metal to its ultimate strength Su,
dicular to the stress, in this case the circumferential thickness profile, as and this is compounded by a bending stress of 70 percent Su. The fact
illustrated in Fig. 7.4. The CTP is the line joining the points of minimum that the rupture was a violent and large opening of the pipe wall,
wall measurements around the circumference. rather than a leak, is addressed and explained in Chap. 13.
The longitudinal stress due to bending and internai pressure is
7.15 Process Pipe Fitness-for-Service-
PD
aL = - + .
075.M
l-Z Level 1
4t
A 14-in carbon steel pipe (ASTM A 106 Grade B) with 0.375-in standard
where M = bending moment in service due to weight and expansion, wall has a design pressure of 800 psi at a design temperature of 100°F.
in·lb The original design corrosion allowance was 0.10 in. After five years of
Z = section modulus of the pipe, approximated by A X t operation the wall thickness is measured by straight beam UT, with
A = cross-sectional area of pipe opening, in2 readings taken at Y.-in intervals along the length of the pipe. The read-
ings were recorded as 0.375 in, -D.375 in, -D.200 in, -D.200 in, -0.150 in,
If, on the basis of a pipe flexibility analysis, the bending moment due -D.200 in, and -D.375 in. There has clearly been corrosion. The cause of
to expansion and weight at the ruptured section is M = 10,000 ft·lb, corrosion has been eliminated, and the future corrosion allowance is
then with the original wall thickness of 0.39 in, the total longitudinal FCA = O in, at least until an outage in seven months when the pipe will
stress, equal to the longitudinal pressure stress plus the stress due to be replaced. Can the pipe be left in service for the next seven months?
bending, would be The minimum wall thickness required by ASME B31.3 piping code is
PD +0.15i M = 130 x22 + 1 x 10,000x12 t . = 800 psi x 14 in = 0.275 in

4t 1t(D2 I 4)t81 4 X 0.39 1t(222 I 4) X 0.39 mm 2(20,000 psi X 1 + 800 psi X 0.4)
= 1800 + 800 = 2600 psi
The minimum measured wall thickness is
where M = bending moment in service due to weight and expansion,
tmm =0.15 in
in·lb
Z = section modulus of the pipe, approximated by A X t
A = cross-sectional area of pipe opening, in2 R= 0.15 in -O =0.545
t 81 = thickness required for supplementary (bending) loads, in I 0.275in
2
This stress of 2600 psi is quite low, and does not pose a design problem 1 0 545
for a new pipe with a nominal wall. However, when the wall is thinned Q=1.123 ( - · ) -1=0.646
1-0.545 I 0.9
down to 0.024 in and there is no sound metal around to reinforce the
corroded zone, the longitudinal stress becomes Length of thickness averaging
aL = PD +0.75i~ L= QJDi = 0.646-.114 in X 0.275 in = 1.27 in

4tmm zmm
22 10 000 12 Averaging the wall thickness readings over 1.27 in around the lowest
= 130x +1x • X J 30,000+13,000= 43,000
4 X 0.024 1t(222 I 4) X 0.024 reading leads to an average wall measuring tam = 0.225 in.
Condition 1, Level1: then the defect will not fail i f the pipeline is pressurized up to a pres-
sure equal to
tam - FCA = 0.225 in - O in < 0.275 in
The measured wall thinning does not pass Levell; Condition 1, Levei 2 2 05
P = Sy t = 60 000 x · = 3000 psi
2 will be checked. D ' 20
Condition 1, Level2:
ASME B31G applies under the following conditions:
0.225 in - O in < 0.90 X 0.275 in
• The metalloss contains no cracks.
The measured wall thinning does not pass Level 2. The system should • The metalloss may be externai or internai.
be derated to operate at a lower pressure or repaired.
• The metalloss could be caused by corrosion or by grinding of a defect.
• The metal loss has a smooth contour, with no stress risers.
7.16 ASME B31 G for Pipelines: What ls lt?
• There are no dents or gouges.
ASME B31G3 is a well-established method to evaluate the integrity of
hydrocarbon pipelines (crude oil, liquid products, or gas), subject to wall • The material is ductile.
thinning. The method is based on initial theoretical work by Folias in • The pipeline integrity is governed by pressure hoop stresses, with
the 1960s,7 followed by testing and further refinement by J. Kiefner insignificant contribution from bending, shear, or tension.
and others in the 1970s.~>-Io Tests of remaining strength of corroded • The rules apply to straight pipeline sections and long bends.
pipe can be conducted on actual corroded sections or on machined
defects such as shown in Fig. 7.12.
Before proceeding further, it is important to understand practically the 7.17 Basis of ASME 831G
B31G criterion. Acceptance of wall thinning in accordance with ASME
Consider a corroded cylinder with a local thinned area, as indicated in
B31G signifies that the thinned wall would be able to withstand a
Fig. 7.13. Ifthe corrosion profile is mostly longitudinal, with a length
hydrostatic test at a pressure that would bring the nominal wall right
not longer than
to yield. In other words, if we have a 20-in diameter and 0.50-in wall
pipeline, API 5L X60, with a defect oflength L and depth d, and ifthis
L <·J20xDt
defect is calculated to be acceptable in accordance with ASME B31G,
then ductile failure of the corroded ligament occurs when the nominal
stress S (stress away from the defect) reaches the following limit
(Folias' equation)7
S -S l-AI A 0
p - nowl-(AIA )IM
o
where Sp = hoop stress at ductile fracture, psi

Sflow = flow stress of the material, psi
A = cross-sectional area of metallost, in2
Ao = nominal cross-sectional area, in2 = Lt
M = Folias bulge factor
M = ) l + 0.8L2
Figure 7.12 Machined longitudinal defect. Dt
Thinning 273
272 Chapter Seven
L
I
I I
I I
: r-d :
~~~--~----------~- -- ------~--------------~
t A d
-· --·-----·-----·-----·-------- --------·--
Figure 7.13 Corrosion profile.
2 3 4
Note that the above limit on corroded length L< '.'20 x Dt can be writ-
ten in a different form, the one adopted by ASME B31G, as
u,[Dt
Figure 7.14 B31G acceptance curve.
Because Sy cancels and M is a function of the defect L, we obtain the

relationship between the defect depth d and the defect length L,
If the corroded section is long (A> 4), we conservatively assume that
the whole wall, all around, is uniformly corroded to a depth d, and the 2
condition for ductile failure becomes L=l.l2Jfii dlt -1
( l.l(d I t)- 0.15 )
S - PD -S
corroded - 2(t _ d) - tlow In conclusion, given a defect of depth d and length L, if L is shorter
than the above value, the corroded pipe can sustain a hydrotest ata
If we impose three more conditions. we will find the B31G equation: pressure that would cause a hoop stress PD/(2t) equal to yield in the
nominal uncorroded section. These types of short defects are permit-
• The first condition is to assume that the flaw shape is parabolic, so ted in B31G.
that the lost metal area is It is standard practice to present pig inspection readings on an
ASME B31G chart where the normalized defect depth (d/t) is plotted
A=~Ld as a function of a normalized defect length Ll...flfi (Fig. 7.14). Results
3 are presented based on this relationship. When wallloss d has reached
80 percent of the nominal wall, the corroded section is rejected, hence
• The second condition is to select a material flow stress Snow equal to
the cut-off at dlt = 0.80.
l.lSy.
• The third and last condition is the expectation that the defect should
resist without failure a hydrotest in which the hoop stress in the 7.18 Derating a Pipeline
sound metal reaches yield. A pipeline that does not meet B31G can be repaired or derated to
If we substitute these three conditions into the Folias equation, we get operate at a pressure P ' lower than the original design pressure PD
where P' is given by3
PhydroD =S =llS 1-%d lt
2t y . y %d /t
1---
M
7.19 B31G Example for Gas Pipeline In the modified B31G method, the failure stress is8•9
(-PD) .
A 14-in X 0.5-in wall, API 5L X40 gas pipeline is procured with frac-
1 -0.85(d 't>
ture toughness requirement in accordance with ASME B31.8. The . =(S}.+ 10ks1 )1-0.85(d/t 1/M)
pipeline exhibits a region ofwallloss, 20-in long and 0.2-in deep. There 2t fa1lure
are no cracks, dents, or gouges. 2
The Folias bulging factor is modified to be for L 1(Dt)::;; 50,
1. Can B31G be applied?
2
2. Is the defect acceptable?
3. What is the maximum pressure at which the line can operate? J
Lz Lz
M = 1 + 0.6275 Dt -0.003375 Dt ()
Solution
1. Yes, the conditions ofapplicability ofB31G, Sec.7.16, are met. L2
M = 0.032- + 3.3
2. For the defect to be acceptable, the defect length must be less than Dt
2 In the NG-18 method. the failure stress is10

04
L = 1.12.J14 X 0.5 ( · ) -1 = 2.8 in
1.1 x0.4 -0.15
PD) _2Sflow 1 -(d/tX1
(2t failure
1-d /t
/ M)
1t
cos (e-Y)
Because the defect is 20 in long, it is longer than L = 2.8 in, and
therefore it does not pass ASME B31G. The pressure in the pipeline
should be reduced to a derated pressure.
y = CCVN>En
2
4AL(Sn0 " )
3. The derated pressure is
P' = l.1x1000 1 - 23 (0.4) =838 psi

1- 2 3 (0.4)[11 ~1 +(0.8x 202 )/ (14 x .os]
2
M = 1 + 0.6275 ~ - 0.003375 2 (
~
2 )
Therefore the line can be operated at 838 psi.
7.20 Modified B31G

L2
The acceptance criteria of B31G can be refined, and made less conser- M = 0.032 Dt + 3.3
vative, by introducing the following variations:
• The use of a flow stress larger than 1.1Sy, such as the mean of Sy where CVN = Charpy V-notch toughness at minimum operating tem-
andS,. perature, in·lb
E = Young's modulus of pipe material, psi
. . 2
• The extension of the B31G curve even where L > -J20 x Dt A = cross-sectional area of Charpy spec1men, m
• A different more precise shape than the parabola and therefore a dif-
ferent value of A than 2/3 dL In the API 579 method, the failure stress is1
These variations will raise the acceptance curve, and permit more wall
loss than the original ASME B31G.
PD)
(2t failure
1 -d/t
=Suo" 1-(d/tXl / M )
276 Chapter Seven Thlnnlng 2n
where RSF -= remaining strength factor

FLD = failure load of damaged component, lb or in-lb
FLU = failure load ofundamaged component, lb or in·lb.
In the case of API 579, the flow stress depends on the remaining In the case of ASME B31G, the failure load of the damaged compo-
strength factor (margin) selected. nent is equal to or larger than yield in the pipeline, so that the lowest
In the British Gas-DNV method, the failure stress is 11•12 value of FLD is
PD)
(2t _S 1-d / t FLD = Sy
Caílun - c 1- (d l t X11 M )
The failure load of the undamaged component is chosen to be the flow
stress S ,, which was selected to be l.lSy,
M=J1+0.31;
FLU = l.lSy
In the 1999 Batte11e method, the failure stress is 13 The remaining strength factor in B31G is therefore
RSF = ~-=0.91
1.1Sy
X - 0.222L If we define the flow stress as Sy 10 ksi. then, for Grade B carbon
- Jn~t-d steel:
35
RSF=---= 0.78
7.21 The RSTRENG• Method 35+10
The U.S. Department ofTransportation (DOT) has approved RSTRENG~ For a higher strength steel such as API 5L X60,
as an alternate method to ASME B31G. In RSTRENG$ the corrosion
profile is refined to closely match the actual corroded area, compared to 60
RSF = = 0.86
the parabolic shape assumed in B31G. The RSTRENG® method is 60+10
available on user-friendly software, and results m less conservative
results than B31G.I4
For example, a 30-in diameter X 0.375-in wall API 5L X52 pipeline 7.23 Steam Condensate
operating at 940 psi has developed externai pitting, up to 37 percent A 3-in sch. 40, ASME B31.1 steam condensate drain line operates at
through the wall. The degraded condition is evaluated using B31G atmospheric prer--sure. The line has a design pressure of 40 psi. This
and RSTRENG®. Both methods conclude that the pipeline needs to design pressure was established assuming accidental plugging of the line.
be derated to operate ata lower pressure, but whereas B31G limits In normal service, the condensate drains by gravity flow at the bottom of
the operating pressure to 650 psi, RSTRENG® permits oper ation at the pipe, following the pipe slope. The line was in service for twenty years
890 psi. before it developed pinhole leaks.
The leaking section was replaced by another schedule 40 pipe, and
7.22 The Remaining Strength Factor in 831G new inspections were conducted after ten more years of service, indicat-
ing continued corrosion with a 60 percent wall loss at the bottom of the
The remaining strength factor is defined as pipe (Fig. 7.15). The minimum measured wall thickness is tmm = 609(
x 0.216 m = 0.13 in.
RSF= FLD The three conditions of a Level1 assessment can readily be evaluated.
FLU The minimum wall thickness required by ASME B31.1 is
tam- FCA = 0.13 in - 0.01 in = 0.12 in
is larger than
tmin = 0.005 in
The corroded pipe, including a future corrosion allowance of0.01 in for

one year of service, is not in danger of bursting.
Condition 2, Level I. The minimum measured thickness tmm must
not be thinner than 0.10 in. This criterion is intended to prevent pin-
hole leaks:
tmm- FCA = 0.13 in - 0.01 in> 0.10 in
This condition is met. The 60 percent corroded pipe section is not in

danger of pinhole leaks for one more year of service.
Condition 3, Leveli. The minimum measured thick.ness tmm must
not be thinner than half the original wall thickness. This criterion is
Figure 7.15 Wall thinning at bottom of condensate tine. intended to flag out significant wallloss. This condition is met beca use
tmm - FCA = 0.13 in- 0.01 in> 0.216 in/2 = 0.11 in.
t = PD ~~-:--4_0_x_3_.5_ _ :--:- = 0.005 in In conclusion: the corroded pipe sections comply with Levei 1 crite-
nun 2(SE + Py) 2(15, 000 X 1.0 + 40 X 0.4)
ria, for one more year of service, based on a constant corrosion rate of
To this thickness, we add a corrosion allowance and the 12.5 percent 10 mils/yr.
mill tolerance from the ASTM material specification:
7.24 The ASME VIII, Div.1 , App.32 Method
t = (0.005 in + 0.10 in) X 1.125 = 0.12 in
A 0.25-in-deep and 6-in-long gouge on a sphere will be repaired by
Only a 0.12-in wall is needed but a schedule 40 pipe, wall thickness 0.216 grinding to a smooth contour, leaving a local thinned area (LTA). The
in was used because the material was available. The wall loss from gouge is 10 ft away from a welded attachment. The ground-out area
0.216 in down to 0.13 in, in ten years, indicates that the corrosion rate will have a final profile as shown in Fig. 7.16, with Rb = 2 in. The LTA
is much higher than estimated at the design stage. will be evaluated in accordance with the rules of ASME VIII, Div.1,
It is not necessary to calculate the length of thickness averaging L
because the axial extent of corrosion S is large, as evident in Fig. 7 .15.
Therefore
tam = tmm = 0.13 in
The corrosion rate over the past ten years has been
0.216 in-0.13 in= 0.009 in/ yr = 9 mils/yr (mpy)

lOyr
Condition I, Level I. If we wish to operate for one more year, the Figure 7.16 Planned profile of the
repaired area.
expected future corrosion allowance would be 0.009 in- 0.01 in; the first
fitness-for-service condition is met for the next year because
Appendix 32 Local Thin Areas in Cylindrical Shells and in Spherical

Segments of Shells.l5 Lmsd ~ 2.5 JRi -7 120 ~ 2.5 ~292.8 X 2.34 = 65 met
The vessel parameters are:
Rb~ 2d -7 2:;::: 2 X 0.25 met
Material= ASME II SA-414, Grade G
Lb;;::: 3d -7 1.0:;::: 3 X 0.25 met
ID = 585.6 in, sphere inner diameter
t = 2.5 in, nominal wall thickness One condition is not met. The thickness of the remaining wall is insuf-
ficient. Weld deposition after grinding to rebuild the wall or other repair
E = 1.0, weld joint efficiency factor
techniques that restare the wall thickness are necessary.
MAWP = 400 psi
Corrosion allowance = 1/16 in
7.25 The ASME XI Code Case N-480 Method
The minimum wall thickness required by ASME VIII, Div.l is
This Code Case applies to the evaluation ofwall thinning dueto erosion-
corrosion in single-phase water conveying piping system.s in nuclear
_ 400 X 292.7 = .
t. - 2 .34 m power plants. According to ASME XI Code Case N-480, the piping must
mm 2x 25,000x1-0.2x 200 be repaired, replaced, or evaluated for acceptability for continued service
According to Appendix 32 the ground-out area on the sphere surface when
must meet the following six conditions.
87.5 percent foom > tmm > 30 percent tnom
where tmm = minimum measured thickness, in, and tnom = nominal
pipe wall, in.
When the extent of wall thinning below tmio is

tmin -tL S3/16 in
L < 2.65 ~Rtmin
where L = length of wall thinning region below tmin, in
R = radius, in
tmin = minimum code required wall thickness, in
In addition, if t > 1.13 tmin, then, the measured wall thickness taloc is
acceptable if it meets the following two conditions.
where tL = thickness of the remaining wall, after grind and blend, in
tmin = minimum wall thickness required by design code, in
DL = length of local thin area in spherical segment, in
R = sphere radius, in taloc >- tmm. [1.5.JRi::::.(1-
L t . )+1]
toom
mm
Rb = radius of blend curvature at ends of LTA, in
Lb = length of blend transition area, in
d = depth of local thin area, in
fL ~ 0.9 tmin -7 2.5 in- 0.25 in = 2.25 in~ 0.9 X 2.34 in met
= 2.11 in
7.26 Widespread Pitting
DLS ~Rtmin -7 6 S ~292.8 X 2.34 = 26 in met
A common forro of pitting is broad, widespread pitting that affects a
lmm - fL ::; 3/16 in -7 2.34- 2.25 = 0.09 in ::; 0.19 in notmet large area of metal. In these common cases, the pitted region should
Figure 7.17 Selective corrosion at seam weld.2

Figure 7.19 Local pitting readily characterized.
E
Figure 7.18 Pitting treated as local thin area. o
..;'
be treated as general corrosion. The pitted zone is evaluated as if it were

shaved out (Figs. 7.17 and 7 .18) and what remains is evaluated using
Figure 7.20 Pitting parameters.
the wall thinning rules.
7.27 Localized Pitti ng 7.28 Example Pitting in Pipeline
When pitting corrosion occurs as a limited number of pits (a dozen or As a result of holidays in the coating, pits have developed at the outer
fewer) that can be clearly characterized for depth, diameter, and pitch, surface of a 20-in pipeline. The line was uncovered and the pits' depth,
a special procedure can be applied, which is more realistic than assum- diameter, and distance were characterized. The evaluation offitness-for-
ing a full region of wall loss. This type of localized pitting occurs, for service will proceed step by step, following the rules of API 579.1 The pit
e~ample, at coating holidays in buried pipes or tanks, as illustrated in parameters are illustrated in Fig. 7.20: the depth (w), the depth relative
F1g. 7.19. to tnun (w'), the diameter (d), and the pitch between pits (P).
Pitting in ductile materiais results in weeping leakage, rarely in
rupture (burst). The risk is that the pits would propagate into cracks. Step 1. Design and fitness-for-service parameters.
There are pitting rules in API 510,16 but ifthey are not met there is no
OD 20 in X w.t. 0.25 in pipeline API 5L X60 ERW
rerating guidance, whereas API 579 will give rerating rules. The evalu-
ation of pitting corrosion in API 579 is based on the ligament efficiency P = 1000 psi
rules of ASME VIII, Div.1, UG-52.15 A limit of 80 percent on pit depth D = inside diameter of pipeline 20 in - 2 X 0.25 in = 19.5 in
is the same consensus number as ASME B31G. A Levei 3 approach
FCA =O in
would be based on fmite element analysis, but this is rarely done for
pitting degradation. RSFa = 0.90
284 Chapter Seven Thlnning 285
Step 2. Pit size and distance. Example: Pit 1 couple 1
d = diameter of pit, reported in tines 7 and 8 of Table 7.1 wl.1 = 0.05 in - (0.25 in - 0.23 in - 0.0 in) = 0.03 in
P = pit-couple spacing (pitch), reported in line 10 ofTable 7.1
Step 5. Average depth and pitch, lines 9 and 10 of Table 7 .1.
w = depth of pit, reported in lines 2 and 3 of Table 7.1
d = dik +djk = 0.20+0.15 =o 18 .
Step 3. Calculate the Code minimum required wall tm1n. 8\"g,k 2 2 . m
Minimum required wall thickness (ASME B31.4),

d =.!. ~ d = 0.05+0.05+0.05+0.15+0.15 =0 10.
8'1: n~ 8vg.k 5 . m
PD 1000x20
t . = = = 0.23 in
mm 2(0.72SyE) 2x0.72x60,000xl
p =.!. ~ P,k = 1.5+2.0+1.0+2.0+1.5 =1.6 in
8 vg n L. 5
Step 4. Calculate pit depth relative to tmin and tabulate results for
each pit couple, as in Table 7.1. Step 6. Remaining strength factor.
Depth ofpits relative to tmin• wik = Wik- (t- tm.n- FCA), reported in
lines 4 and 5 of Table 7 .1. p8vg -d8vg 1.6-0.1 - O
•94
11 - -
r8vg- - -
p 8"g 1.6
TABLE 7.1 Parameters for Pit Assessment, in
Couple Couple Couple Couple Couple

1 2 3 4 5 Average
1 - Pits in couple 1- 2 3-4 5-6 7--8 9-10 w~,-g E 8 ,-g (t - FCA + W~vg- tmin)
- 1--- +--
RSF - " - - - - -----=c."-------=c.:::.__
2 - Depth of odd pit. w 0.05 0.05 0.08 0.10 0.10 tmin tmin
3 - Depth of even pit, w 0.05 0.05 0.08 0.15 0.15
RSF = 1 _ 0.07 + 0.81(0.25- O+ 0.07- 0.23) = l.0
4 - Depth odd relative 0.03 0.03 0.06 0.08 0.08 13
lmm,w' 0.23 0.23
5- Depth even relative 0.03 0.03 0.06 0.13 0.13
tmin• w
But if the calculated RSF is over 1.0, as is the case here with RSF =
0.11 0.11 0.07
1.013, then RSF = 1.0 should be used.
6 -Average !ines 4 and 5, 0.03 0.03 0.06
W~vg
7 - Width of odd pit, d 0.05 0.05 0.05 0.15 0.20
Step 7. The remaining strength factor. IfRSF :::>: RSFa proceed to Step 8,
otherwise go to Levei 2 or evaluate the pitted zone as a general metal
8 - Width of even pit, d 0.05 0.05 0.05 0.15 0.15
loss, as illustrated in Fig. 7.18.
9 - Average !ines 7 and 8, 0.05 0.05 0.05 0.15 0.18 0.10
d8vg
Step 8. The ligament strength. Calculate R 1 for each pit, reported in
10- Pitcll of couple, P 1.5 2.0 1.0 2.0 1.5 1.6
lines 11 and 12.
11 - Rt odd pit 0.87 0.87 0.74 0.65 0.65
12- R 1 even pit 0.87 0.87 0.74 0.43 0 .43 Rt = tmin - w~ - FCA = 0.23- 0.03- O = 0 _87 max
Line 4 = odd pit relative to minimum wall. lmin 0.23
Line 5 = even pit relative to minimum wall.
Line 6 = average of odd and even pit depths. R = 0.23-0.13-0 = 043 .
Average ofLine 6 = w~,•. t 0.23 · mm
Calculate Q and check ligament strength. For the lowest R, which tred < lmm < lgreen
corresponds to the even pit in couple 5, we obtain where
1 -R~
2 = tmin + FCA
f1-10.43/0.90
2 lgn, n
0 3
Q=1.123\lr ) -1=1.123 - .4 ) -1=0.50
1- R1 RSFa • Green zone, continued operation till next inspection, if
The ligament strength is acceptable if the pit diameter is smaller than
a minimum limit given by
with the next inspection scheduled at a time T;nspect such that
~ = 0.5J20 in x 0.23 in = 1.07 in
Q vu'min
Since the pit width of the even pit in couple 5 is 0.15 in, which is less
than 1.07 in, then this criterion is met. This is the strength check of where FCA = future corrosion allowance, not to exceed tmm - lycllow, in,
the pitted metal surface. and CR = corroswn rate based inspection results, in/yr.
The acceptance of pit depth, to prevent pin bole leaks, is checked by
the condition Rr ~ 0.20. In this case Rr minimum is 0.43 and therefore the For example, a pressure vessel has the following parameters:
condition is al~o met. In this exercise we have checked one coupling of 14
pits into 7 couples. In practice, the assessment must be repeated for each tnom = initial nominal wall thickness = 0.50 in
pit combined \\ith another pair to form a new couple until ali permuta- tmín = minimum wall thickness required by ASME VIII, Div.1 = 0.30 in
tions of adjacent pits are exhausted. This is a time-consuming process.
Currently under development for API 579 is a Leve] 1 pitting assessment tmm = minimum measured wall = 0.20 in
procedure in which the pattern of pitting is compared to standard pitting CR = corrosion rate estimated to be 10 mpy = 0.010 in/yr
charts, in a manner similar to the current acceptance criteria for rounded FCA = future corrosion allowance for a period of five years =5 X
indications in weld radiographies, ASME VIII, Div.1, Appendix 4. 0.010 m = 0.05 in
7.29 Simple Criterion

The red zone corresponds to a wall thickness below tred given by
In light of all of the above, and in the absence of regulatory or other

mandatory requirements, a simple practical criterion for fitness-for- t red = max { -0.30
3
-; 010.
. m; -0.50}
-
2
=max(o.10m;
. o.10m; 0.25 m)
. · 0.25 in
service assessment of wall thinning may be as follows:
• Red zone, immediate shutdo\vn, assessment for fitness-for-service Because lmm = 0.20 in is thinner than tred the defect should be imme-
CAPI 579, ASME VIII, Div.l Appendix 32, ASME XI. or ASME B31G diately assessed for fitness-for-service or repaired. Note, for informa-
where applicable), or repair if tion, that in this case the green zone corresponds to a measured wall
thickness above tgreen given by
lgreen = 0.30 in + 0.05 in = 0.35 in.

where
References
t red- 3 ' .
tnom)
- max( tmm. o 10 in·
' 2
1 API RP 579, Fitness-{or-Sen.:ice, American Petroleum Institute, Washington,
DC.
2. Kiefner & Associates, Worthington OH
• Yellow zone, assessment for fitness-for-service within 30 days (API 3. ASME B31G, Manual for Determining the Remaimng Strength o(Corroded Ptpelilles,
579, ASME VIII, Div.1 Appendix 32, ASME XI, or ASME B31G as Supplement to ASME B31 code for pressure piping, 1991, American Society of
applicable if
288 Chapter Seven
Chapter
8
4. API Standard 650, Welded Steel Tanks for Oil Storage, American Petroleum Institute,
Washington, DC.
5. API 653, Tank Inspection, Repair, Alteration, and Reconstruction, American Petroleum
6. ASME B31.1, Power Piping, American Society of Mechanical Engineers, New York.
7. Folias, E. S., An Axial Crack in a Pressurized Cylindrical Shell, lnt. Journal of
Fracture Mechanics, Vol.l, No. 1, 1965.
8. Kiefner, J . F., Vieth, P. H.,A Modified Criterion for Evaluating the Remaining Strength
of Corroded Pipe, AGA Project PR3-805, AGA catalog No. L51609, December 22, 1989.
9. Kiefner, J . F.. Vieth, P. H., New method corrects criterion for evaluating corroded pipe,
Oil & Gas Journal, August 1990.
Geometric Defects
10. Kiefner, J. F., Maxey, W. A., Eiber, R. J., Duffy, A. R., Failure Stress Leveis ofFlaws in
Pressurized Cylinders, ASTM STP536, 1978.
11. O:NV Recommended Practice RP-FlOl, Corroded Pipelines, Det Norske Veritas, 1999.
12. Kirkwood, M .. Bin, F., Improved Guidance for Assessing the Integrity of Corroded
Pipelines, ASME Pressure ~ssel and Piping Conference, Hawaii, 1995.
13. Stephens, D. R., Bubenik, T. A., Francini, R. B., Residual Strength ofPipeline Corrosion
Defects Under Combined Pressure and Axial Loads, AGA NDG-18 Report 216,
February 1995.
14. Technical Toolboxes Inc., RSTRENCJ3, Houston.
15. ASME Boiler and Pressure Vessel Code, Section VIII, Pressure ~ssels, American
Society of Mechanical Engineers, New York. 8.1 lntegrity of Geometric Defects
16. API 510, Pressure Vessel Inspection Code, Maintenance, lnspection, Rating, Repatr,
and Alteration. American Petroleum Institute, Washington, DC. Geometric defects are unintended distortions in shape. They include gen-
eral shell distortion and flat spots, out-of-roundness, ovality, bulging,
blisters, weld misalignment, dents (local depression, Fig. 8.1), and
gouges Omifelike cut, Fig. 8.2).
8.2 Assessment Steps

When a cylindrical shell is subject to pressure, hoop stresses develop
in the shell wall (Chap. 3)
These stresses are constant through the wall; they are labeled mem-
brane stresses (Chap. 3). Ifthe shell is deformed by a geometric defect,
then, in addition to the hoop stresses crh, bending stresses crb will also
develop through the wall (Fig. 8.3). The evaluation ofthe deformed shell
for fitness-for-service consists of the following steps.
• Calculate the hoop membrane stress.
• Calculate the bending stress that appears as a result of the deformation.
• Add the hoop membrane to the bending stress.
• Calculate the fatigue life in cyclic service subject to the calculated
hoop plus bending stress.
289
290 Chapter Elght Geometric Defects 291
Figure 8.3 Hoop crA causes bending cr6 at discontinuity.
Figure 8.1 Dent in wal1. 1
Figure 8.4 Profile of corroded bottom course, with radial bulging.
8.3 Distorted Tank Bottom Example

A storage tank with the same nominal dimension and characteristics as
the tank in Secs. 7.8 and 7.9 exhibits uniform wall thinning of 0.05 in
all around the bottom course. In addition, the tank bottom bulges out
in the thinned section by 0.20 in, all around its circumference (Fig. 8.4).
Evaluate the fitness-for-service of the tank, with a future corrosion
allowance of 0.01 in.
Figure 8.2 Gouge with crack. 1 Because the corrosion is uniform, the average measured wall is
292 Chapter Eight Geometric Defects 293
tam = 0.55 in- 0.05 in = 0.50 in

By similarity to the evaluation in Sec. 7.10, the first fitness-for-ser-
vice condition is met because
tam- FCA = 0.50 in- 0.01 in > tmin.650 = 0.40 in.
The second and third fitness-for-service conditions are also met because
lmm- FCA = 0.50 in- 0.01 in> max (0.5 tnom; 0.10 in)= 0.275 in.
To complete the fitness-for-service assessment, the outward bulging
must also be evaluated. A finite element model of the tank, with the thin
wall, and fixed boundary condition at the tank base, subject to the hydro-
static pressure of contents, shows that the maximum stress occurs at the
shell-bottom plate junction, point A in Fig. 8.5, where the stress intensi-
ties are calculated and classified in accordance with Chap. 3, as mem-
brane, bending, and peak stresses. The r ange of stress intensity between Figure 8.6 Bent risers.
a tank filled to its maximum height (maximum stress) and empty (min-
imum stress) is calculated and divided by two to obtain the altemating
stress S..Jt· The altemating stress Sa1t is entered into the fatigue curve
(Chap. 3) to determine the number of cycles that can be sustained.
8.4 Accidentally Bent Riser Example

A boat accidentally impacts 4-in diameter risers on the side of an offshore
platform (Figs. 8.6 and 8.7). The risers did not leak. Can they be kept in
service?
To solve this question, we ask ourselves: if pipes are commonly cold
bent in the shop or in the field; why is this any different? The solution
~. ••. -. -- I
I -------:...-
Figure 8.7 Riser viewed from platform.
Figure 8.5 Simplified tank model.
to this problem is to assess the conditions for shop or field cold bending
of line pipe, and to compare these controlled bending conditions to what
happened to the riser. The requirements for field cold bending of liquid
pipelines are ASME B31.4 code:
• The minimum radius of bends for NPS 12 and smaller pipe shall be
at least 18D, where D is the diameter.
• The wall thickness after bending shall not be less than the minimum
permitted by the pipe specification.
• Bends shall preserve the cross-sectional shape of the pipe (no dents,
see Sec. 8.5, no cross-sectional distortion).
• Bends shall be free from buckling, cracks, or other evidence of
mechanical damage.
• The pipe diameter shall not be reduced at any point by more than
2.5 percent of the nominal diameter.
• The bend shall be able to pass the specified sizing pig.
• Tangents approximately 6 ft in length are prefe1Ted on both ends of
cold bends.
The following additional conditions also apply in the case of the
damaged risers. Figure 8.8 Pipe bent after beavy snow.
• Verify that the impact did not cause a gouge or crack in the pipe wall.
• Nothing. This is usually the case for dents in shells with small Dlt,
• Verify that the impact did not damage the riser coating and jacket. for example, a dent in a 4-in sch. 40 pipe Wlt = 4.5/0.237 - 20).
• Verify that the impact did not fail riser supports or guides. • An immediate puncture causing a leak or rupture.
• Verify that the impact did not cause a leak of the platform flange; • A delayed rupture. This is the case for unconstrained dents in rela-
torque check flange bolts. tively large D/t shells, operating ata high hoop stress (on the order
• Verify, by PT, MT, RT, or UT that the large plastic deformation ofthe of one-third of yield), and subject to load fluctuations. For example,
platform bends did not cause cracks. a dent in a storage tank with frequent changes in liquid levei, or a
dent in a large liquid pipeline with continuous pump-induced pres-
These conditions, and any other system-specific concerns, can be sure fluctuations.
checked on the accidentally bent riser and, ifthey are ali met, then the
riser may be left in service. An unconstrained dent acts as a membrane that is free to breathe in
The pipe in Fig. 8.8 was permanently deformed under large snow and out as the pressure fluctuates (Fig. 8.10). This is typically what
loads. The same approach can be applied to evaluate the fitness-for- causes the dent to eventually fail by cyclic fatigue. A constrained dent
service of the accidentally bent pipe. cannot breath in and out. This is the case, for example, with a dent at
the bottom of a pipeline resting on a rock protrusion. The rock bearing
8.5 Dents in Pipelines against the pipe wall does not allow the dent to breathe out as the
pressure increases.
A dent is a local inward distortion of a shell wall (Figs. 8.1 and 8.9). lt A dent in a liquid hydrocarbon pipeline (oil and gas) has to be eval-
usually happens as a result of an impact or a concentrated bearing force uated according to regulations that stipulate repair under the following
on relatively thin shell (large D/t). There are three outcomes to dents: conditions. 2
/
I
/
Ho I
~
I
I
I
I
I
t
Figure 8.11 Dent with gouge.

Figure 8.9 Internai view showing dents. 1
_..,
./ dent), but right where the impact occur red there may also be a cut
./
./ (a gouge) .
./
./
./ Four problems arise at this gouge:
./
./
./ • A concentration of stresses
./
/ • Possibly a crack, Fig. 8. 2 and 8.11
/
/
/ • A microstructural change due to the heat generated at the point of
/ impact
/
/. Figure 8.10 Dent breathes as • A deposition of backhoe metal onto the dented component
pressure fluctuates.
These four problems complicate the fitness-for-service assessment to the
point where it may be safer and more effective to repair the damage.
The European Pipeline Research Group (EPRG) has developed a
• The dent is in a seam or girth weld. technique to assess the failure pressure and the fatigue life of pipelines
• The dent contains a scratch, gouge, or groove. with dents and gouges, under the following conditions:3
• The pipe operates above 40 percent Sy and the dent depth in NPS 12 • Pipe size 6 in to 42 in
and larger pipe exceeds 0.25 in; or ifthe dent depth exceeds 2 percent • Wall 0.23 in to 0.64 in
nominal diameter in pipe larger than NPS 12.
• Yield stress Sy of 36 ksi to 70 ksi
• Charpy toughness 2/3 specimen llJ to 142J
8.6 Dents with Gouges
• Dent depth over diameter (unpressurized) 0.25 percent to 10.15 percent
Things get more complicated if a dent also includes a gouge (deep
knife-edge cut in the surface; Fig. 8.11). For example, if a backhoe • Gouge depth over thickness 1.4 percent to 66.3 percent
impacts the side of a tank there will be a distortion of the shell (a • Gouge length 2 in to 21.5 in
298 Chapter Elght Geometric Oefects 299
8.7 Wrinkles and Buckles K = curvature of bent line, l /in

Wrinkles and buckles in pipelines can happen as a result of soil move- ó = midspan deflection of bent líne, in
ment, pipe relocation, offshore vessel impact, field bending, handling, L = span of bent tine, in
lifting, or lowering, and so on (Figs. 8.12 to 8.14 '· . T = wall thickness, in
To prevent buckling, the strain in bent pipe must be below buckhng
For example, if a 20-in OD X 0.5-in wall pipeline settles 6 in over a
limit strain.4 '6
length of 100 ft, the strain in the compressive side of the settlement
bend is
E= DK = 4D.D < 2.4(!...)1.6
2 2
L D E-
_ 4 X 6 X 20 = O N
2 . 04 -1(
1200
where E = compressive strain in buckle
D = diamete~ in The strain at buckling is
16
= 2.4 ( o· )
5 ·
Eb = 0.66%
20
Figure 8.12 Pipeline buckle dueto soil movement.
Agure 8.13 Cracked buckle in a high-pressure pipeline. 1 Figure 8.14 Buckled pipeline offshore. 1
300 Chapter Elght
Geometric Defects 301
3 -. - - - + - - - - - _,_ -
I I
I I I I
--;- - -~--- - - - - - T- - - - - -~- - -- - -
1
2 .. - _._-----
o~
I
.J
----------- -.-. - ---- -

I
10 20 30 40 50 60
oh, ksi
Figure 8.15 Operating limits for npples in pipelines.6 Figure 8.16 Lamination and blisters.
In this case E < Eb- As a rule of thumb, if a pipe operates above 30 per-
cent Sy any wrinkle should be removed. If the pipe operates below 30
percent Sy the wrinkle could be treated as a dent.
8.8 Mild Ripples

A ripple is a mild wrinkle (waviness) on the surface of bent pipes and
shells. Recent research on the integrity of ripples in pipelines indicates
that they can cause a potential for leak if the ripple is large Oarge ripple
height over diameter) and the pipeline operates at high hoop stress.
Recently, criteria have been proposed to limit the ripple height to pipe
diameter ratio d/D as a function of operating hoop stress (Fig. 8.15).6
8.9 Blisters
Blisters can forro in vessel, tank, and pipe walls by the accumulation
of hydrogen at imperfections (such as laminations or inclusions), and
in banded microstructures. At these imperfections atomic hydrogen Figure 8.17 C'rack in blistered plate. 1
combines into larger H 2 gas molecules, eventually causing bulging
(blistering) and at times cracking (Figs. 8.16 and 8.17).
• Component operates below the API 941 curves for hydrogen service.'
8.1 o Fitness-for-Service of Equipment • Material remains ductile, with no loss of toughness.
with Blisters
~e first step is to map the blisters, as shown in Fig. 8.18.
API 5798 provides a set of evaluation cri teria for the fitness-for-service . ~hsters may be acceptable, without repair, under the following con-
assessment of blistered steel under the following conditions. ditions.
Figure 8.18 Blister parameters.
• They are small (less than 2 in diameter).

• They do not project beyond 10 percent oftheir diameter (Bp s 0.1 C). Figure 8.19 Peaking assessment
parameters ( not to scale).
• They contain no crack (the assessment does not apply to me or
SOHIC, Chap. 5).
• They are at least 2t from welds and 2-{jji from structural discontinu-
ities.
Example, Step 1. Data:
• The thinnest measured ligament tmm is larger than half of the nominal
wall. Pipe OD = 36 in
Wall t = 0.5 in
8.11 Assessment of Weld Misalignment
and Shell Distortions
R= 17.50 in inside radius
Material = 1-1/4Cr 1'2Mo steel with
Ali tanks, vessels, and piping have a certain degree of out-of-roundness
shell distortion), but these should have been kept within the toler- E = 25.5 X 106 psi
ances of the construction code or the engineering design. For example, v = 0.3 Poisson ratio
the fabrication tolerance of API flat bottom storage tanks are:
Design pressure = 315 psi
• Out of plumb by no more than 1/200 of the total height Design temperature = 800°F.
• Out of roundness less than 1 percent, within 1 ft frorn the bottom S a = 16800 psi allowable stress at 800°F.
• Out of roundness within ~ in on radius if the diameter is below 40 Joint efficiency = 100 percent (E = 1.0)
ft, and Y. in if it is larger than 40 ft
FCA = 0.05 in
• Weld offset misalignment of x6 in if the tank wall is less than I, in thick,
and ~ in if the tank wall is thicker than ~~ in Peaking distortion ô = 0.20 in
If the distortion exceeds the construction tolerance, it becomes neces- Step 2. Mernbrane stress:
sary to analyze the condition for fi.tness-for-service, by fi.rst calculating
the hoop stress and second calculating the bending stress induced by the P(R+FCA )
hoop stress applied to the deformed shape, as described in the following crm - E t - FCA +0.6
sections.
- 315psi ( 17.50 in+ 0.05 in o6) 12 .
8.12 Fitness-for-Service Assessment - 1.0 0.5 in - 0.05 in + · - •500p8l
of Peaking
Peaking is a form of shell distortion, where a weld peaks outward (Fig. Step 3. Ratio of the mduced bending over the applied membrane
8.19). stress:
12(1- v2 )PR3
Sp = E(t-FCA)3
2 3
= 12(1-0.3 )315(17.5+0.05) =2.8
(25.5 X 106 X0.5- 0.05)3
c, 0.5
ô 0.20
0.011
R 17.50+0.05
Enter Fig. 8.208 with Sp - 2.8 and 8IR = 0.011 --+ Cr - 0.90.
The r a tio of induced bending stress to applied membrane stress for the
longitudinal joint of a cylinder with angular misalignment is
o 10 20 30
g!Ja = 6Ô C = 6 X 0.2 in 0.90 = 2.4
b t-FCA r 0.5 in-0.05 in Sp
Figure 8.20 General form of peaking coefficient Cr for 'NR • 0.01.

This means that the pressure causes in the metal wall: (a) a membrane
(uniform tensile) stress O'm = 12,500 psi, and (b) because ofthe peak at
the weld, a bending stress equal to 2.4 times the rnembrane stress, 2.4
x O'm = 2.4 x 12,500. Therefore, the total stress at the peaked weld is
Longitudinal stress = a2 = 12,500 I 2 = 6250 psi.
12,500 psi x (1 + 2.4). Radial stress = a 3 = O.
Step 4. Remaining strength factor. Step 3. The stre~s intensity is the largest difference between principal
With a factor Hr = 3.0 because the induced stress is secondary we stresses
calculate the remaining strength factor as,8
t:.Srange = max(42,500-6250 ; 6250-0; 0-42,500) = 42,500 psi
RSF= HrSa = 3x16,800psi = 1.19 And, because we will be using a fatigue assessment curve based on
a (l+Rb) 12,500psi x(1+2.4)
testing of smooth bar specimen made of base metal, not welds, we
apply a stress concentration factor of K 1 = 1.5 to the stress intensity8
Step 5. Compare the calculated remaining strength factor at the weld
RSF = 1.19 to the allowable remaining strength factor RSFa = 0.90. K !:!.Sra....,.., = 1.5 X 42,500 = 63,750 psi
In this case, the peaking effect is acceptable because 1.19 > 0.90.
Fatigue assessment with peaking. Can the pipe cycle 200 times O to Step 4. The alternating stress is Satt = 63,750 I 2 = 31,875 psi = 32
315 psi? ksi; the corresponding number of fatigue cycles is approximately
20,000 cycles, based on ASME Boiler and Pressure Vessel Code design
Step 1. Fatigue assessment. fatigue curves for smooth bar specimens (Fig. 8.21). Note, as explaincd
in Sec.3.32, that Fig. 8.21 is a design curve, with safety factors.
Evaluate peaking at 315 psi without cycling, as was done in Step 4
above.
8.13 Deformed and Repaired
Step 2. Determine the total stress range. Vessel Explosion
Circumferential stress, membrane plus bending dueto peaking = cr1 The catastrophíc explosion of a 7.5-ft diameter x 8-ft tall stainless steel
= 12,500 psi + 12,500 x 2.4 = 42,500 psi. vessel illustrates the need for careful repairs of geometric defects. The
accident had all the signs of an overpressure rupture: an accidental over-
Geometric Defects 307
306 Chapter Eight
1e+1 1e+2 1e+3 1e+4

Number of cycles Figure 8.23 Vessel hcad found 100 yards away. 10
Figure 8.21 Fatigue curves for carbon steel.
words, the explosion pressure did not exceed 130 psi, and the vessel, if it
was not darnaged, should haYe been able to sustain even 180 psi. The
accident investigation noted that the vessel had been weakened when it
"had been deforrned twice due to misapplication of vacuum... "
Differential externa! pressure due to vacuum could cause local or general
inward buckling of a vessel or tank, which would be a classic example of
geometric defect. The damage is reported to have been repaired, but "the
repairs were not mspected or certified to rneetASME Code requirernents."
8.14 Defects Beyond Assessment

In rnany cases, geometnc damage is beyond assessment, and the com-
ponent needs to be repaired, as illustrated in Fig. 8.24 where an acci-
dental impact caused the distortion of flange and bolts.
References
Figure 8.22 Area of vessel ex.plosion.
10 1. Kiefner & Associates, Worthington, OH.
2. Code of Federal Regulations, Title 49 'Iransportation, Part 192, 'Iransportation o(
Natural Gas and Other Gas by Pipeline: Mimmum Federal Safety; Part 193, Lique(ied
heating of the contents, no relief device, a violent longitudinal rupture Natural Gas Facilities: Federal Safety Standards; Part 194, Response Piam; for
Om;hore Oi/ Pipelines; Part 195. 'Iransportation of llazardous Liquids Pipelines.
along the shell and separation of the head at the head-shell weld seam 3. Roovers. P.. et ai., Metlwds for Assessing the To/erance and Resistance o( P1peline11 to
(Figs. 8.22 and 8.23). But what appears more diffi.cult to e~pl~ is the Externai Damagc, 1999, EPRG, European Pipeline Research Group.
fact that experts' calculations indicated that the vessel, m tts newly 4. Stephen, D. R., et ai., Pipeline .Uonitoring · Limit State Criteria, ~G-18, AGA, 1991.
constructed condition, should have been able to sustain 180 psi, yet "it is 5. Antaki, G. A., Píping and Pipeline Engineering, Dekker, New York.
6. Rosenfeld, M., et al.. Development of acceptance cri teria for mild ripples in pipeline
improbable that the pressure in the vessel exceeded 130 psi...."' In other field bends, in International Pipeline Conferencc, September 2002. ASME. Alberta,
Canada.
308 Chapter Eight
Chapter
9
Cracks
Figure 8.24 Damage beyond assessment.

9.1 Cracklike Flaws
7. API RP 941. Steels for Hydrogen Servtce at Elevated Thmperatures and Pressures
in Petroleum Refinenes and Petrochemical Plants, American Petroleum lnstitute, Cracklike flaws are single or multiple narrow, sharp cracks in a compo-
Washington. DC nent; they can be superficial or embedded. They can preexist in the base
8. API RP 579, Fitness-for-Service, American Petroleum Insti~ute, Wash.ington, DC. metal, orbe created during fabrication (e.g., a narrow lack of penetra-
9. AS.ME Boiler and Pressure Vessel Code, Section ll Matertals, Amencan Soctety of
Mechanical Engineers, New York. tion or a shrinkage crack in a weld), or they can appear in service, for
10. U.S. Chemical Safety and Hazard Investigation Board, lnvestigation Report example, as a result of corrosion cracking (Fig. 9.1) or fatigue (Fig. 9.2),
Cata:;trophi~ Vessel Failure, D.D. Williamson & Co., Inc., Aprilll, 2003, Report No.
or at impact gouges (Fig. 9.3).
2003-11-I·KY, ~arch 2004.
Heavy pitting can also result in a through-wall crack (Fig. 9.4), and
individual narrow pits can also behave like cracks, for example, if the
radius at the base of the pit (Fig. 9.5) rbase is such that1
r bese < max (0.25 tmln; 0.25 in).
Figure 9.1 Corrosion cracking along weld HAZ.
309
31 O Chapter Nine Cracks 311
Figure 9.4 Crack caused by externai corrosion.2
Figura 9.2 Crack from mechanical

fatigue.
Figure 9.5 ::->arrow pit behaves as crack.
First, the brittle fracture prevention condition: a parameter called

the stress mtensity, calculated at the crack, must be below the fracture
toughness of the metal3
K<Kc
where K = stress intensity, ksi..Jin (Secs. 9.3 to 9.5), and Kc = fracture
toughness of the metal at operating temperature, ksivin <Sec. 9.6).
Second, the ductile tearing prevention condition: the stress in the lig-
ament of metal remaining behind the crack must be smaller than the flow
Figure 9.3 Crack at impact gouge.
stress ofthe metal cr1;g <Sr. where Ot;g = ligament stress (Sec. 9.8), ksi and
Sr = flow stress ofthe metal at operating temperature, ksi CSec. 9.9).
9.2 Crack Stability 9.3 Stress lntensity

A crack in metal subject to tensile str ess is stable (i.e., it will not t~nd The stress mtensity applied to the crack is a parameter that depends
to t ear ) if two conditions ar e met: the first condition prot ects agamst first on the direction of the applied load relative to the cr ack (Fig. 9.6).
brittle fracture of th e cr ack , and the second protects against ductile If the applied load tends to tear open the crack in tension, the stress
tearing of the crack. intensity 1:> labeled "mode r and written as K 1 If the applied load
312 Chapter Nine Cracks 313
11 111
i
~
!
f F=..--------- f .~ -f
Figure 9.6 Three modes of fracture.
tends to open the crack in shear, in-plane, it is Ku, and out-of-plane it

is K 111, as illustrated in Fig. 9.6
The most common mode of failure of cracks in tanks, vessels, and
pipes is mode I, and we therefore focus on the mode I stress intensity
Kz. The stress intensity has the general forro
Figure 9.7 Oil transm1ssion pipeline.
where K 1 = mode I stress intensity, ksi"'in

cr = applied tensile stress at the crack, ksi
a = crack size, in
F = geometry factor
In this formula, the applied stress cr is the stress at the point of crack-
ing calculated as if there were no crack. The geometry factor dependa
on the shape of the component where there is a crack, and the relative
location of the crack in the component.
9.4 Example-Crack in Pipeline

The oil pipeline shown in Fig. 9.7 is 36-in OD x 0.50-in wt (wall thick-
ness), API 5L X42 carbon steel. The maximum allowable operating
pressure (MAOP) is 600 psi at 120°F maximum operating metal tem-
perature. It had been in service for over 20 years when it developed
externai longitudinal cracks along the pipe seam weld, as shown in Fig. Figure 9.8 Crack along IongJtudinallieam.
9.8. The crack depth was measured by shear wave (angle beam) ultra-
sonic inspection; the deepest crack was 0.20 in deep.
The applied stress direction that tends to open the longitudinal crack Because there are multiple, long, aligned cracks (Fig. 9.8), we con-
is the hoop (circumferential) stress. The magnitude of the hoop stress, servatively assume that the crack 1s infinitely long. In this case, the
as if there were no crack is parameter a in the formula for stress intensity K is the crack depth a
= 0.20 in.
cr = PD = 600x36 = 21 600 si The geometry factor F is a function of the shape of the crack and the
2t 2x0.50 ' p shape of the component that contains the crack. For an infinitely long
crack on the outer diameter of a cylinder under internai pressure, the
cr = 21.6 ksi geometr) factor F depends on the following ratios. 4
2a = 2x0.20 =0.4 For example, to apply the API 579 method to estimate fracture
Dout - Din 36- 35 toughness for a 0.5-in-thick API 5L pipeline, we first determine the
reference temperature Tref for the material and thickness, Fig. 2.34, as
Dirt = 35 =0.97 T rer = -15°F. Then, we calculate the fracture toughness Kw and the
arrest fracture toughness (lower bound, dynamic toughness) KIR: 1
Dout 36
Ktc = 33.2 + 2.806 exp[0.02(T - Tref + 100)]
For these parameters F = 2.0. Therefore, the stress intensity for the
crack in Fig. 9.8 is KIR = 26.8 + 1.223 exp[0.0144(T - T ref + 160)]
K 1 = 21.6 x .Jn x o.2 x 2.0 = 34.2 ksirm where T = lowest operating temperature, °F, and T rer = reference tem-
perature of the material at thickness, °F.
Note that we only addressed the pressure-induced hoop stress
If the lowest operating temperature is, for example, 40°F, then
applied to the crack. In the case oftbis pipeline, because the cracks are
near the longitudinal seam, we must also consider the weld residual K 1c = 33.2 + 2.806 exp[0.02(40 - (-15) + 100)] = 95 ksi--Jin
stresses. The question of residual stresses is addressed later in this
chapter. KIR = 26.8 + 1.223 exp[0.0144(40 - ( - 15) + 160)] = 54 ksi--Jin.
9.5 Stress lntensity Solutions 9.7 Weld Residual Stresses
The stress intensity solutions K1 for different shapes of cracks and In the vicinity of welds, weld residual stresses have an important influ-
components can be obtained from fracture mechanics textbooks or ence on the stability of cracks. Weld residual stresses vary through the
stress intensity solution compendiums. 1•4 •5 thickness, and they can vary from tensile to compressive within the same
weld. The magnitude of weld residual stresses depends on the material,
the weld procedure (in particular, the heat input), the degree to which
9.6 Fracture Toughness the welded parts are restrained, the wall thickness, and the type of post-
weld heat treatment, if any. Weld residual stresses can be measured and
Fracture toughness K 1c is a measure of the ability of a material to sus-
predicted by analysis, as described in Chap. 4. .
tain a stable crack under tensile stresses. The fracture toughness of a
For the purpose of fracture mechanics, the stress intensity K is the
material is established as described in Chap. 2, either by direct testing,
sum of two contributions: a stress intensity K due to stresses from
or by approximation based on the Charpy V-notch toughness.
applied loads such as pressure, weight, or temperature, and a stress
In the example of the pipeline of Fig. 9.7, the API 5L line pipe steel
intensity K due to residual stresses. We illustrate this point through
was fabricated over 20 years ago, and there are no records of its frac- an example later in the chapter.
ture toughness Kw. Line pipe fabricated to today's API 5L, starting
with the 42nd edition of API 5L, is designated by a product specifica-
tion levei 1 or 2 (PSL 1 or PSL 2). A PSL 2 API 5L X42 line pipe will 9.8 Ligament Reference Stress
have as a minimum a Charpy V-notch toughness of 20 ft·lb trans- The stress in the remaining ligament beneath the crack (ali8 ), also
verse and 30 ft-lb longitudinal at 32°F. In practice, modern line pipe called the reference stress, can be readily calculated for simple geome-
steel is procured with much higher toughness. When the toughness is tries. In the simplest case, a cylindrical shell of thickness t contains an
unknown, it should be measured on a specimen of the actual material, infinitely long crack of depth a (Fig. 9.9).
particularly in environments that can adversely affect toughness The remaining ligament has a thickness (t - a) and the stress per-
over time, such as aging of Cr-Mo steels, ferritic stainless steels, and pendicular to the crack, the hoop stress, in the remaining ligament is
Duplex stainless steels, operating in the 700 to l100°F range, or
hydrogen-embrittled steels. A lower-bound toughness may be estimated PD
as described in API 579. 1 crlig = 2(t- a)
9.9 Flow Stress

The flow stress is a material strength property. We know that a metal
yields at the yield stress S and ruptures shortly after reaching the ulti-
mate strength S~. Somewhere between these two points, the material
has lost its practical engineering strength; it "flows" and is no longer fit
for service. The point between S.v and Su where the metal starts to flow
is the flow stress. It has been defined differently by various authors and
Figure 9.9 Long crack of depth a. applied differently in various standards. Some common definitions of
the flow stress are
For our earlier pipeline example, Fig. 9.7, the ligament stress is
600x36 . where Sr = flow stress, psi

cr1ig = 2 x (O.S-0.2) =36,000 ps1 Sy = yield stress, psi
S u = ultimate strength, psi.
(Jbg = 36 ksi
S = S, t 10,000 psi S r= 1.1 x S.v
Because the ligament in a ductile material tends to bulge before rup-
For the API 5L X42 carbon steel pipeline of Fig. 9.7, the flow stress
ture, the exact ligament (reference) stress is more complex than this
may be defined using any of the above formulas; for example,
simple approximation; it is a function of a shell parameter, similar to
the Folias parameter in Chap. 7, S.- 1.1 X 42 ksi = 46.2 ksi
9.1 O Foundation of Fracture Assessment,

the FAD
where À. = shell parameter The foundation of fracture assessment of cracklike fla ws resides in the
c = crack length parameter, in two conditions presented in Sec. 9.2:
R; = inner radius of shell, in
a = crack depth, in
For through-wall cracks, where the remaining ligament is zero, the Rather than verifying these t wo conditions separately, crack stabil-
reference stress takes a different forro; for example, for a long through- ity is determined as an interaction between the ratio K, = K 11Krc and
wall crack in a cylindrical shell, the reference stress is1 the ratio Lr = rrii/Sr, or in some cases, such as API 579, L, = rrug!S,
where
• K, is a measure ofthe brittle fracture potential at the crack.

• Lr is a measure of the risk of ductile (tearing) fracture at the crack.
M 1 = ~1 + 0.4845Ã. 2
The curve that plots the limiting value of Kr against Ln the interac-
tion curve, is called the failure assessment diagram or FAD, presented
wherePb = primary bending stress at the crack, ksi (Chap. 3), and Pm = in Fig. 9.10. The exact FAD must be obtained from the applicable code
primary membrane stress at the crack, ksi (Chap. 3). or standard.1 The equation K,.{L,) of the FAD curve is
Krc = 54 ksi"in
(Jlig = 36 ksi
0.8 ----------~----L-
1 ----~-----------~----------
Sr = 46.2 ksi
1 I
I
I Therefore
LEFM I X• I '
0.6 ----------~L----------~----------- ~ ----------

1 : K ,. = Kr = 34.2 = 0.63
I K 1c 54
I
I EPFM I I
0.4 -------~--------------~--
I
L - crlig -~-
I I
--------~----------
1 I
,.- s, -46.2-0.78
I
I
I ,,..,.,,.,.... I I
0.2 - - -I- - - - - - - - - - - -- ..,..._ - - - - J.. - - - - - - - - - - -

I ~~~ I I
Sr =46.2 ksi
I _.....-
/ ,.. ..............
I _.....- LL The point X in Fig. 9.10, falls below the FAD diagram line. At this
!,....,.._,
0~---------r---------+--~----~~----L_~
moment in time, the crack, subject to the operating pressure, is mechan-
o 0.5 A 1.5 8 2 ically stable. This does not mean that it may not continue to grow; the
degradation mechanism that caused the crack to appear in the first
Lr
place may well continue and drive the crack deeper. As the crack depth
Figure 9.10 Failure assessment diagram FAD.6 progresses, a increases, and K 1 and <TJig increase, until one of two out-
comes takes place:
• A leak develops if the crack depth equals the wall thickness and
point Xis still under the FAD line, or
Cracks with L, and K, below the curve are stable cracks; above the • Point X reaches the FAD line; at this point the crack becomes unstable
curve the crack is unstable and causes a fracture. The AA limit in Fig. and the line will rupture.
9.10 applies to C-Mn steels, and the BB curve applies to austenitic
This section il1ustrated the foundations of fracture mechanics. For
stainless steels. Also, note the three regions:
clarity, the method has been stripped ofits complexity but, as described
• LEFM is the zone of linear elastic fracture. In this region, above the so far, it is insufficient to make a reliable decision. Next, we assess
FAD curve, failure would occur by brittle fracture. another crack, to illustrate the method in its full, but necessary, com-
plexity.
• EPFM is the zone of elastic plastic fracture. In this region, above the
FAD curve, failure would be a mixed elastic and plastic fracture.
9.11 The 15 Steps ot Crack Analysis
• LL is the zone of limit load fracture. In this region, above the FAD
curve, failure would be by ductile fracture ofthe remaining ligament Step 1. Description
behind the crack. Step 2. Stress distributíon at crack
Step 3. Material properties
The FAD is an extremely clear and resourceful way of assessing
Step 4. Crack síze
cracklike flaws. For example, in the case of the cracked pipeline (Fig.
9.7), we have calculated Step 5. Partial safety factors
Step 6. Reference (ligament) stress
Kr = 34.2 ksi"in Step 7. The L, ratio
Step 8. Stress intensity dueto Pm Step 3. Material Properties.

Step 9. Reference stress due to secondary and residual stress Yield stress at design temperature Sy-= 38 ksi
Step 10. Secondary and residual stress reduction factor
Ultimate strength at design temperature = 70 ksi
Step 11. Stress intensity due to secondary and residual stresses
Flow stress
Step 12. Plasticity interaction factor
Step 13. The K, ratio
Step 14. FAD assessment
s, ;o: Sy + s. . ;o: 54 ksi
2
Step 15. Crack instability limit
Young's modulus of steel E = 29.7 106 psi
9.12 Vessel Example NDT of 0.75-in thick SA 516 Gr.70 material at minimum tempera-
ture of 70"'F, T ..,r = -41°F
Step 1. Description.
Material toughness is obtained using API 579 Appendix F. For a
Pressure vessel, ASME VIII Div.1, 1999 metal temperature as low as 0°F, the fracture toughness is
Material = ASME BPV Code Section II, SA 516 Grade 70
K~e = 32.2 ._ 2.806 exp[0.02(T - T rf'f + 100)]
OD = 120 in, thickness = O. 75 in wall
No postweld heat treatment K 1c = 32.2 2.806 exp[0.02(0 - (-41) + 100)] - 80 ksi..Jin
MAWP design pressure) = 125 psi Using this approximation, K1c cannot be larger than an upper bound
T max = design temperature =- 120"F of 100 ksi'-lin, which is the case here because KIC = 80 ksi'iin.
T min = minimum operating temperature = 70"'F Step 4. Crack size.
Ste p 2.Stress distribution at crack. Crack in shell, away from structural discontinuities
The crack is longitudinal; the stress perpendicular to the crack is the Axial crack parallel to a longitudinal seam of the shell, in the heat-
hoop stress due to pressure affected zone
p ;o: PD ;o: 125x120 ;o: 10 000 si Crack length - 6 in
m 2t 2x0.75 ' p Crack depth = 0.20 in
The crack is near a longitudinal V-groove seam weld, with no post- Step 5. Partial safety factors.
weld heat treatment; in accordance with API 579 Appendix E, the Accepting a probability of failure of 10·3 , in other words acceptmg a
residual stress distribution perpendicular to a longitudinal seam weld, 1/1000 chance that the calculation shows the crack to be stable when
from the inner diameter (x = O) to the outer diameter (x = t) is a poly- it actually is not, for the material and conditions selected, API 579
nomial stress distribution 1 Chap. 9 provides a partia} safety factor on applied stress PSFs = 1.5.
The toughness is a lower-bound estimate, so that th ere will be no
~
2
additional partial safety factor on K, PSFK = 1.0. The crack is
cr'(x)= 0 0 +01 (f )+cr 2( ) assumed to be very well characterized, so that there is no partial safe-
;o: 48 -1.536 X 102 X ( ~) + 1.536 X 102 X ( r

7 ksi
ty factor on crack size PSFa = 1.0; in summary: PSFs = 1.5, PSFK
1.0, and PSFa = 1.0.
Because PSFS is 1.5, the calculation will proceed on the basis of a
maximum hoop stress equal to P m = 1.5 X 10,000 psi = 15,000 psi =
where cr0 = Sy -r 10 ksi = 48 ksi. 15 ksi.
Step 6. Reference (ligament) stress. The stress intensity due to the polynomial distribution of residual
The reference stress for primary loads is stresses is obtained from a "K solution" textbook4•5 or API 579
Appendix C, 1 and calculated to be KR = 10.8 ksi..JIIi.
<Tref = MtPm = 1.07 X 15,000 = 16,100 psi
Step 12. Plasticity interaction factor.
Step 7. The L, ratio. The plasticity interaction factor is a factor that accounts for the effect of
L, is defined in this case relative to the yield stress Sy rather than the plasticity at the crack tip on the stress intensity due to secondary and
flow stress s,.
is residual stresses. It depends on the ratio Lr due to primary stresses and
L~ due to secondary and residual stresses. For primary stresses, Lr =
Lr == crrer = 16,100 = 0.42 0.42 was calculated in Step 7.
Sy 38,000 The ratio L~R for secondary and residual stresses is
Step 8. Stress intensity due to Pm· L SR == 24.3 == 0.64

r 38
The stress intensity factor dueto the primary membrane stress caused
by internai pressure is obtained from a "K solution" textbook4 •5 or API Given L; = 0.42 and L~R = 0.64, we obtain from API 579, Chap. 9,
579 Appendix C, 1 factors 'I'= 0.0476 and cp = 0.5541. With L~R $ 4.0, we obtain
K == Pm x.fMxF=3.5 ksirm
cp == 1.0 +"' == 1.0 + 0.0476 == 1.086
cp 0.5541
Step 9. Reference stress due to secondary and residual stress.
There are no imposed secondary stresses, and the residual stress is the Step 13. The Kr ratio.
polynomial stress distribution calculated in Step 2. The corresponding The stress intensity ratio is
reference stress is calculated based on API 579 Appendix D to be
Kr = K ~ cpKSR = 3.5 + 1.086 X 10.8 = 0 _19
cr~f == 24.3 ksi KIC 80
Step 10. Secondary and residual stress reduction factor. Step 14. FAD assessment.
If secondary and residual stresses cause the ligament to be plastic (u:; In conclusion, the 6-in-Jong, 0.20-in-deep surface crack in the vessel
+ u;:: > Sy), then the stresses tend to relax, which is accounted for by a shell, near the longitudinal weld, ata pressure of 125 psi, can be rep-
secondary stress reduction factor. But if(u:; + u~) < Sy there is no sec- resented by the following point on the failure assessment diagram:
ondary and residual stress reduction factor. In this example, there are no
secondary stresses (displacement controlled and thermal stresses) u;:! <Lr; K.) = (0.42; 0.19)
= O, so we only consider u ::;. There is no secondary stress reduction This point is plotted as point A in Fig. 9.11. lt is well within the stable
because
region of crack stability. If the crack does not grow, it will not rip through
cr ~; = 24.3 ksi < Sy = 38 ksi. the wall. More interesting is to assess at what point the crack may
reach the FAD line; this is done in Step 15, Sec. 9.13.
Step 11 . Stress intensity due to secondary and residual stresses.
Because there are no secondary stresses, the stress intensity due to 9.13 Margln to Failure
secondary stresses K 5 p]us residual stresses K R is
The above procedure can be continued one more step, Step 15, to pre-
KSR = Ks + KR = O + KR dict at what depth or length a crack becomes unstable and will cause a
9.14 Leak Through Cracks

The crack opening area of a crack in a shell subject to internai pres-
sure is
0.8 I
----T.- --------~----------
I
1
COA = 2~tacrc2
I E'
I
0.6- ----- ----/ - ,8-----~- where COA= crack opening area, in2
Kr
I I a = membrane tensile stress applied to crack, psi
I
I I
I c = half length of through-wall crack, in
- - -·- - - - - - - -- --
I
0.4 -- 1-----~- -- E' = Young modulus in plane strain, psi

I I
I I
I I
I
AI
.... ! ____
0.2+ f ..... ------.... --~
I
---
I I
I ....
I
For a longitudinal crack in a cylinder
I LL
....
o •' .... a = 1 + O.lÀ. + 0.16Ã?
o 0 .5 1.5 2
Lr
Figure 9.11 Result plotted as a FAD point.
For a circumferential crack in a cylinder

TABLE 9.1 FAD Point for a Growlng Crack
Crack Crnck
depth length,
a, in 2c, in K, L À. 1.818c
0.2 6.0 0.19 0.42 JR.i
0.30 9.0 0.25 0.46
0.40 12.0 0.32 0.51
For example, in the case of a vessel with an operating pressure of
747 psi. a diameter D = 54 in, anda thickness t = 0.63 in, the hoop
0.50 15.0 0.42 0.57
stress is
0.60 18.0 0.61 0.66
= PD =
747 x 54 = 32 014 psi 32,000 psi
crh. P 2t 2x0.63 '
fracture. In this step, we repeat the above FAD analysis by progres-
sively increasing the crack length and depth, until we reach the FAD If the through-wall crack length is 4 in, then c 2 in and
line that indicates fracture. In the vessel example above this procedure
leads to the results of Table 9.1. À.= 1.818x2 =0.9
The progression of (L,; Kr) is tracked in Fig. 9.11 from point A to .j(54/2)x0.63
point B. The crack, even having progressed to being 18-in long and
0.60/0.75 = 80 percent through-wall remains stable. a= 1 +0.1 0.9 +0.16x 0.92 = 1.22
Wíth a Young modulus in plane straín E' of 28 106 psi, the crack open-
ing areais
2
COA= 2xl.22 .;32,000x7tx2 = _ in
6 0 035 2
28x10
The flow rate of a liquid through an orifice (the crack) is given by
where cd = díscharge coefficient of orifice

COA ;;;; crack opening area, ffl
g = gravity, ft/s2 32.2 ft/s2 Figure 9.12 Stable, narrow circumferential crack.
jJ> = pressure across the crack, psi
3
p = liquid density, lb/ft
For this example
:lP = 747 psi

p = 62.4 lb/ft3
cd = o.61
COA = 0.035 in2 = 2.4 x 10 4
rt2
therefore
2 32 2 44 747
q=0.61x2.4 x 10 4 x · xl x =0.049 ft_J / s
62.4
= 2.94 ft 3 /min 22 gal/min
9.15 Application of Fracture

Mechanics to Fatigue Figure 9.13 Unstable crack ran along seam weld.
Once a fatigue crack has been initiated, fracture mechanics can be used
to predict its growth rate through the thickness, and to assess whether References
the crack will be stable (leak; Fig. 9.12) or unstable Cbreak; Fig. 9.13). 1. API RP 579, Fitm!~-ç.for·SI'roice. American Pctroleum Institute, Washington DC.
The fracture mechanics approach to fatigue analysis is addressed in 2. Kiefner & A!>sociaU>s, Worthingt{)n, OH.
3. Barsom, J . M., Ro fe, S. T. Fracture and Fat1gue Control in Structures ~~. We~~
Chap. 3. Conshohocken. PA.
328 Chapter Nine
Chapter
10
4. Rooke, D. P., Cartwright, D. J., Compendium o{Stress lntensity Factors, UKProcurement
Executive, Ministry of Defence
5. Tada, H., Paris, P. C., lrwin, G. R., The Stress Analysis of Cracks Handbook, ASME,
NewYork.
6. Scott, P. M., et ai., Review ofExisting Fitness-For-Service Criteria for Crack-Like Flaws,
WRC Bulletin 430, April 1998.
Creep Damage
10.1 What ls Creep?

When metais are held at temperatures above approx:imately X of their
melting point (e.g., 2800°F/3 - 900°F for steels) they undergo three
damaging phenomena, referred to as creep damage:
• A continuous increase of strain (deformation) under constant operat-
ing load (Figs. 10.1, 10.2, and 10.6).
• A deterioration of the microstructure, with formation of voids along
the grain boundaries, leading to fissuring as the voids grow and con-
nect with each other (Fig. 10.3). The macroscopic aspect of advanced
creep is illustrated in Fig. 10.4.
• A gradual spheroidization and precipitation of carbides at the grain
boundaries in ferritic-pearlitic steels (Fig. 10.5).
Creep curves such as those in Fig. 10.1 can be obtained from manu-
facturers of alloys for high-temperature service, and are developed
based on standard creep test procedures.3 Note in Fig. 10.1 the three
stages of creep:
• Initially first-stage strain accumulation is quite rapid; microvoids

are isolated cavities at grain boundaries.
• Then second-stage creep occurs over a longer period of time, with a
low, nearly constant, strain rate. Cavities are oriented and start
microcracks.
• Third-stage creep is where fissures are either visible or detectable by
radiography in the field; at this point failure is imminent.
329
330 ChapterTen Creep Damage 331
Figure 10.1 Steady increase in

strain under constant load.
Figure 10.4 Long-term creep

damage.
t
Figure 10.2 Creep swelling under internai pressure.l
Figure 10.5 Rupture due to graphitization.2
The generally accepted temperatures for onset of creep are obtained

from ASME B&PV Code Section li, Part D, Tables lA and lB, where,
for each material, a note "T" indicates the onset temperature for creep,
and the need to consider "time-dependent properties." An example of
onset creep temperatures for common metais is given in Table 10.1.
10.2 High-Temperature Corrosion

Concurrent with creep, other high-temperature corrosion mechanisms
take place, such as high-temperature oxidation (Fig. 10. 7), discussed in
Figure 10.3 Creep-induced microvoids in steel.
Chap. 5, which further complicate the assessment offitness-for-service
TABLE 10.1 Onset of Creep4
Metal Onset of creep, °F
Carbon steel with Su s; 60 ksi 650

High-strength carbon steel St· > 60 ksi 700
C-~o 750
1XCr-XMo 800
2Y.Cr-1Mo to 9Cr-1Mo 800
Stainless steel 304 950
Stainless steel 316, 321, 347 1000
Alloy 800 and 800HT 1050
Figure 10.7 Thermal oxidation and spalling ofmagnetite.2
Council (MPC) and the American Petroleum Institute (API), have con-
ducted extensive research in creep design and creep failure. The phe-
nomena involved in creep are complex, and do not lend themselves to
Figure 10.6 Creep fatigue rupture ofmain steam pipe. 1 simple assessment methods. It is no surprise that in the first issue of
API 579, the creep assessment chapter, was the only blank chapter, to
be completed "later."5
and remaining life. Table 10.2 presents the approximate temperature for
There are several difficulties in predicting the fitness-for-service and
onset ofscaling, where oxidation equals 10 mg/cm2 in 1000 h (-41 days).
remaining life of components operating in the creep regime, such as
boiler and furnace tubes, or high-temperature steam pipes in fossil
10.3 The Difficulties of Creep Analysis power plants. These difficulties include:
and Predictions
• Obtaining and modeling temperature- and material-specific proper-
The fossil power industry, through expert groups such as the Electric ties, including time-dependent material properties (where strain and
Power Research Institute (EPRI) and the refining and petrochemical temperature govern time to rupture), high-temperature plasticity,
industries, through expert groups such as the Materiais Properties and thennal-stress coupled effects.
TABLE 10.2 Approximate Onset of Scaling2 a small change in temperature, say from 300 to 310°F, would have a
negligible effect on the component's life.
Metal Approximate onset of scaling, oF
Carbon steel with Su 5 60 ksi 1025

10.4 Short- and Long-Term Overheating
High-strength carbon steel Su > 60 ksi 1025
C-}~Mo 1050
Short-term overheating is when the metal temperature significantly
exceeds its design limit for a briefperiod. The mechanical properties of
1Cr-}g\Jo 1100
the metal are reduced from an already low value, and rupture occurs
2XCr-1Mo 1100 under the operating load. A short-term overheating rupture (Fig. 10.9)
18Cr-8Ni 1600 is characterized by:
18Cr-8Ni + Ti 1700
• Signifi.cant plastic deformation, and swelling and bulging if pressurized
27Cr 2000
• Little oxidation of the surface
• No metallographic evidence of creep void coalescence into cracks
Continuous operation temperature variance, °F • Thin-lip, sharp-edged, fishmouth burst if pressurized
-9 Design 9 18 27 36
604-------+-------L-------~----~~----~ A long-term overheating rupture (Fig. 10.10) is characterized by:
I I I
40 ---- ---~-------~---- -~-~----- --~-------
1 I I • Little plastic deformation, and swelling and bulging if pressurized
I I
20 I I I
---~-------~------- ~--- - -- ---------
• Visible oxidation and scale on the surface
1 I I
I • Metallographic evidence of creep void coalescence into cracks
o --- ----- -- r --- ----~-- ----~- ------
1
I
I
I
• Thick-lip, blunt surface burst if pressurized
- 20 ---- L-------~----- --
1 I • Wrinkled surface appearance
I I
I I I
g> -40 -------r-------r -- ----T- ------T------- In some cases, the creep rupture is mixed, starting with long-term
~
1 I
I I
~ - 60 -------~- ----------- ---+ - ----+----- -- overheating, followed by a short-term burst at even higher temperature
o 1 I
I
I I
that causes the rupture (Fig. 10.11).
I I I
- 80 - ------ ,-------.-------~--- ---- ~-------
-1 004-------~------~------~----~~----~
-5 Design 5 10 15 20
Continuous operation temperature variance, oc
Figure 10.8 lnfluence oftemperature variation on creep life.7
• Understanding and applying complex design and evaluation criteria,

based on finite element analysis. s
• Obtaining an accurate temperature and applied load profile, because
creep life is very sensitive to small changes in operating temperature,
and local overheating. This is well illustrated in Fig. 10.8, which
shows that an increase of 10°F in creep operating temperature leads
to a 30 percent reduction in life. This is a very different behavior
than that of metais that operate below the creep regime, where such Figure 10.9 Rupture dueto short-term overheating.2
336 Chapter Ten Creep Damage 337
10.6 ASME 111 NH Method

Ifwe select the creep qualification rules of ASME B&PV Code, Section
III, Subsection NH, Appendix T, then we will have to:
• ~lodel the component with finite elements.
• Enter the inelastic material creep properties.
• Apply the temperature gradient and loads to the model.
• Calculate and combine strains at all points.
• Check three limits on the maximum accumulated inelastic strain:
• One percent for strains averaged through the thickness
• Two percent for linearized strains at the surface
• Five percent for local strains.
Figure 10.10 Long-tcrm crcep rupture with oxides. 2
In addition, ifthe temperature, pressure, or applied load varies with
time, under cyclic stress, the cumulative creep and fatigue damage
interaction shall not exceed a limit D. whích is written as
where D =total creep-fatigue limit per ASME III

n =number of fatigue cycles
Na =number of allowable cycles for a given strain levei for low-
cycle fatigue
M =duration of time interval at a given stress levei
Td =allowable life for a given stress levei for stress-to-
rupture
Figure 10.11 Long-tenn followed by short-tenn overheating. 2
An example ofinteraction limit D is illustrated in Fig. 10.12.
10.5 Creep Assessment Methods

1O.7 Operating Loads
The first step in creep design or fitness-for-service is to select a quali-
fication criterion. There are at least tive possible methods of creep Typically, a component operatmg in the creep regime will be subject to
evaluation: mechanical loads such as pressw·e and weight, and thermal loads such
as expansion-induced bending and longitudinal and through-wall tem-
• The method of ASME III NH (Secs. 10.6 to 10.11) perature proftle. The creep analysis and life prediction are quite sensi-
• The method of API 530 (Sec.10.12) tive to the temperature profile along and within the component, and her e
• The ASME B31.3 method (Sec.l0.14) lies a key difficulty of creep life prediction: an accurate heat transfer
analysis or, better yet, actual temperature readings are essential to accu-
• The Omega-based method of API 570 Appendix 10 (Sec.10.15) rately determine the applied temperature proftle, leading to a mean-
• Strain and metallographic methods tSecs. 10.13 and 10.16 to 10.18) ingful analysis.
338 ChapterTen Creep Oamage 33g
1.0 - r - - - - - - - - -- - - - - ----, e C. crC.
where i = creep strain rate, 1/s

cr = applied stress. psi
cl and c2 =material- and temperature-dependent constants
- Another time-dependent material property is the rupture life of
the material, as a function of time and temperature, as illustrated
- 100
- 80
0.~ -l~~==.=====~;:=====;:::::=J
0.1 1.0
'in
.lC
ui
(/)
60
~
êi5 40
Figure 10.12 lntc>raction limit D for creep assessment.
10.8 Time-lndependent Material Properties 20
Time-independent material properties must be obtained over the range

of operating temperatures, from the room ambient (shutdown) condi-
tion up to the creep (operating) temperature. These properties are: o 2 4 6 8 10
Strain,%
• Modulus of elasticity (Young's modulus). Figure 10.13 Time-independent material propcrtiel;.
• Yield and ultimate strength.
• Coefficient of thermal expansion.
• Elongation at rupture.
• Poisson ratio.
o~
• True stress versus true strain curve. The relationship between stress qS
and strain includes the plastic range; it is nonlinear, therefore, the Cl
c 10° -,--------------
1
stress and strain analysis will also be nonlinear. A bilinear curve may ~
c
be sufficient for analysis (Fig. 10.13). -~
• Fatigue life relationship between cyclic strain and cycles to failure, êi5
as a function o f temperature, in the forro of Fig. 10.14.
10.9 Time-Dependent Material Properties

1~14---------------+-------------~--------------~
The time-dependent material properties define the creep relationship 102
between the creep strain rate (strain increase with time), and the applied Cycles
stress Figure 10.14 Fatigue life cur\'e~.
100~--------~----------,---------~--------~ 100
50 -
111
- I
-----~ ----------- r----------
I
1111
20
111
'(ij
~
CIÍ
111
cn
1+----------r---------+----------r-------~ ~
o 10 100 1000 10,000 Ci5 111
10
Rupture life, h
Figure 10.15 Gi!neral trends in rupture Iife. 111111
in Fig. 10.15. Rupture stress is commonly expressed as a function of

5 I
a parameter that combines operating temperature and time at tem-
perature, the Larson-Miller parameter, as illustrated in Fig. 10.16.
LMP =(T + 460)(C + log t) 10-3
30 32 34 36 38 40
where T temperature, °F LMP
C constant, 20 for ferritic steels, 15 for austenitic steels
Figure 10.16 General tnmd of failure stress as a function of LMP.
t -time at temperature, h
1 0.1 O Creep Life Analysis 0.5
Having established the applied loads, and time-dependent and time-

independent material properties, the creep analysis is typically carried 0.4
out by finite element analysis. First is a heat transfer analysis to accu- ~
o
rately determine the temperature profile along the component and c:
'Cil
through its thickness. I f actual temperature profiles are available, then i5 0.3
the heat transfer analysis may not be required, and the actual readings cn
Cll
should be used. cn
~
The temperature distribution, established by analysis or by actual õ 0.2
readings, is input to the creep stress and strain analysis. The tem- >
perature and concurrent loads are applied to the model, as a func-
tion of time, and the accumulated equivalent strain (von Mises 0.1
strain) is obtained as a function of time, such as illustrated in Fig.
10.17. If the number of cycles is large, for example, 100 cycles, the
strain may be calculated over a smaller number of cycles, for exam-
ple, 10, and extrapolated. This is done for the averaged strain in the o 2 4 6 8 10 12
highest strained section, and the local strain at the highest strained Cycles
point. The objective will be to predict fatigue life (Fig. 10.18). Figure 10.17 Example of equivalent strain versus load cycles.
10.11 Qualification
From the creep analysis, we obtain two results:
• The maximum strain (a) averaged through the thickness, (b) linearized
at the surface, and (c) local. These maximum strains are compared to
the 1, 2, and 5 percent limits, respectively.
• The strain rangeM: and the equivalent stress (von Mises stress) <req
for each cycle. As an example, these strains and stresses are tabulat-
ed in Tables 10.3 and 10.4.
In the example illustrated in Table 10.3, the strains are obtained
from the analysis; the number ofpermitted cycles N are obtained from
the fatigue life curves such as Fig. 10.14. The total cumulated fraction
of fatigue life, the bottom right box of Table 10.3, is the term
In Table 10.4, the equivalent stress is obtained from the analysis, the Figure 10.18 High-temperature thermal fatigue cracks.1
time to rupture at that stress, and temperature is obtained from the rup-
ture life curve, such as Fig. 10.15, and the total fraction "Actual/Rupture,"
10.12 API 530 Creep Assessment
the bottom right box ofTable 10.4 is the term
A furnace exhaust tube has operated for four years at 390 psi and 1650°F
(Fig. 10.19). After four years of service, the insulation was removed,
revealing that the tube had visibly bulged in its top section. Is it safe to
place the tube back into service for another four years?
TABLE 10.3 Strain Contributlon to Creep Fatigue
The tube is 1.66-in outside diameter, 0.25-in wall, ASTM B407 UNS
08811, Incolloy Alloy 800HT. This is an alloy with 30 to 35% nickel, 19
Cycle N 1/N to 23% chromium, 39.5% minimum iron, 0.06 to 0.10% carbon, 0.25 to
0.60% aluminum, 0.25 to 0.60% titanium, and anASTM grain size 5 or
1 0.0071 0.0043 0.0028 20,000 0.00005
coarser. The alloy has a melting range of 2475 to 2525°F.
2 etc. etc. etc. etc. etc. The bulged section has a 1.81-in diameter (a 9 percent increase in
etc. diameter compared to the original nominal diameter of 1.66 in), anda
Total 0.00100 0.22-in thickness (a 12 percent decrease in thickness compared to the
original nominal 0.25-in wall).
Some operators replace tubes when diameter bulging has reached 1
TABLE 10.4 Stress Contribution to Creep Fatigue
to 2 percent for ASTM A297 Grade HK-40 (Fe-25Cr-20Ni-0.4C) and 5
Time slice <Teq. Mpa r.·c Rupt. Td, h Actual .1t, h Actual/rup. to 7 percent for HP modified alloy (Fe-25Cr-35Ni-0.45C-L5Nb). In this
case, the API 530 procedure will be applied to assess the fitness-for-
1 10 660 100.000 0.5 0.000005
service of the tube.
2 7 970 80,000 300 0.00375 A stress analysis of the tube shows that the bending stresses are
3 etc. etc. etc. etc. etc. negligible compared to the pressure stress, because the tube is well sup-
etc. ported and has sufficient flexibility. lfwe assume that bulging occurred
Total
evenly over the four-year period, and divide the period into four slices
0.02000
of one year each, then the fitness-for-service calculation proceeds as
TABLE 10.5 Creep Damage Assessment
Life use.
Time. P. T. t, 00, CJA. Life fraction
Slice }T psi •F in in p,;i L\1P. range range,yr range
1 1 390 1650 0.25 1.66 1295 42.0 42.8 9.1 22.8 0.11 0.04
2 1 390 1650 0.24 1.70 1380 41.4 42.4 4.5 14.4 0.22 0.07
3 1 390 1650 0.23 1.75 1480 41.2 42.2 3.6 8.1 0.28 0.12
4 1 390 1650 0.22 1.81 1600 41.0 42.0 2.9 9.1 0.34 0.11
'lbtal 4 0.95 0.34
Metallography of a section of the removed tube indicated significant

creep voids at grain boundaries, which confinned the decision to remove
the tube.
10.13 Nondestructlve Assessment

In a similar case, furnace tubes also ruptured about four years into ser-
vice. The tubes were ASTM A 297 Grade HK-40. The tubes operated at
1730°F and 310 psi. Litt.le bulgmg, on the order of 1 percent of diameter,
had occurred at creep failure so gauging alone could not help detect
damaged tubes.
Figure 10.19 Fumace tubes. Radiographic testing (RT) was used to examine the tubes in place, but
laboratory tests on samples indicated two shortcomings in applying
presented in Table 10.5, where P is the operating pressure, T is the RT to detect creep fissures:
operating temperature, t is the wall thickness, OD is the outer diameter,
ah is the hoop stress [ah = <PD l1(2t)], LMP is the Larson-Miller param-
• The tubes first had to be drained of their contents to better detect
eter, and life rangeis obtained from API 530.8 the fissures.
With the potential for 95 percent life used, the run-or-repair decision • RT detected third-stage creep cracking when the tube had practically
is made: the risk of creep rupture is too high, the tube will be replaced, exhausted its usefullife, but was not sufficiently sensitive to detect
and all other similar tube:; will be inspected and replaced ífnecessary. earlier stages of creep damage.
The risk is high because
Some tubes with fissures characterized by RT were purposely left
• The consequence of creep rupture in service would seriously set back in place to determine how much longer they would last, and therefore
production. obtaining a method of correlating RT readings with remaining life.
• The líkelihood of creep rupture is significant because up to 0.95 (95 The alternative to RT is ultrasonic inspection (UT), but here too
percentl of the tube life may have already been expanded in the first there were challenges: the rough surface of cast tubes and large grain
four years, and the assumption that bulging occurred evenly may not cast microstructure. But these difficulties were overcome by using a
be true, and actually an accelerated degradation may have occurred dual sensor unit. The attenuatlon signal between the transmitting and
over the latter part ofthe first four years. Also, the 9 percent measured receiving sensors correlated well with the presence and density of
diameter growth is significant compared to the rule of thumb for tube creep fissures, with practically no ultrasound transmission through
retirement. severely fissured areas.
10.14 High-Temperature 831.3 Pipe

Application s = min( Su3 .'1.5'
Sy . SR,mean .
1.5
sR,min) =
' 1.25
min( 55.8 20.4 55 . 35 ) = 13.6 ksi
3 ' 1.5 '1.5 '1.25
A process system operates at 900°F (475°C) and 900 psi (620 MPa). A
surface crack is discovered in a vessel nozzle, and a repair option is The required minimum wall thickness of the repair sleeve is
being investigated to install a repair sleeve around the nozzle. The
sleeve is made ofASTM A 335 Pll 1.Y.Cr-~Mo low-alloy steel. The repair t= 900 X 10.75 = 0.43 in
is to operate for a period of 18 months, until a scheduled outage when 2(13,600 X 0.80 + 900 X 0.4)
the cracked nozzle would be permanently repaired.
In the worst of cases, if the crack were to leak shortly after placing The thickness of the 10-in sch.60 repair sleeve is 0.50 in > 0.43 in,
the repair sleeve, then the sleeve would operate at 900°F and 900 psi for which is adequate for tbe eighteen month service.
18 months ( -13,400 h ). The required wall thickness of the cylindrical
sleeve is calculated using the ASME B31.3 equation for minimum wall
10.15 Draft Method of API 579 Levei 1
thickness
The creep assessment method under development for Chap. 10 of API
PD 5795 will have, like ali other API 579 chapters, three assessment leveis,
t=----
2(SE+Py) from a simple but conservative approach in Levei 1, up to a detailed
analysis in Levei 3. As currently planned, the draft Levei 1 creep assess-
where P =maximum operating pressure, psi ment method in API 579 will predict remaining operating time to reach
D =sleeve outer diameter, in one-quarter of the mean creep life. A 10-in carbon steel pipe, 10. 75-in
S =sleeve material allowable stress, at operating temperature, outer diameter, 0.50-in wall, operates at T = 900°F and P = 900 psi. The
psi nominal stress in the pipe wall is
E =sleeve longitudinal weld joint efficiency
y =code parameter a= PD = 900x10.75 = 9 .7 ksi
2t 2x0.50
In this case, P = 900 psi, D =10.75 in, y = 0.4 (ASME B31.3), andE
= 0.8 (ASME B31.3). We have to determine the allowable stress S. The For carbon steel, at 9.7 ksi- 10 ksi and 900°F, based on the draftAPI
ASME B31.3 code design allowable stress in creep is 579 Chap. 10 curves, the fractional damage rate would be 5 X 10- 5 per
hour. Therefore, the time to reach one-quarter ofthe expected mean life
s = min(Su. Sy. sR,mean. sR,min) under the operating stress and temperature is 1/(5 x 10-5 ) = 20,000 h,
3 '1.5' 1.5 ' 1.25 a little over 2 yr.
where SR,mean = mean creep rupture strength, ksi, and SR,min = minimum 10.16 Life Fraction Analysis
creep rupture strength, ksi. The minimum ultimate strengtb Su and
yield stress Sy ofthe material at 900°F are obtained fromASME B&PV The life fraction analysis applies to predict the remaining life of low-
Code Section II, Appendix D, Tables U and Y-1, as Su = 55.8 ksi and alloy Cr-Mo steels in creep service.9-11 The life fraction remaining LFR
Sy = 20.4 ksi. is given by
The Larson-Miller parameter is E /E
t E •
LFR = 1--= ( 1-- )
LMP = (900 F+ 460) 20+ log 13,400 h= 33 tr Er
1000
where LFR = life fraction remaining
From Figure F.4 of API 530, fifth edition, the mean and minimum t = time in creep service, h
creep rupture strengths ofthe repair sleeve material are 55 and 35 ksi, tr= rupture life, h
respectively. Therefore E = strain at time t
e, = rupture strain 35,000 =

1
ll-[1+11.4x10 ' 6 x(4-1)xt ]1' 11- 4 l)
e. = Monkman-Grant constant, approximately 3 percent for 11.4x10-6 r
low-alloy steel
t, ;;;; 105,495 h = 12 yr
For example, if a component with a rupture strain e,= 10 percent, has
reached a measured creep strain E = 5 percent after 4 yr, t = 35.000 h, If another tube has only thinned down in the same four years
its life fraction remaining is (35,000 h) to 0.20-in, then
10 13
LFR = ( 1- 1~) . = 0.10 K'=
0 25 0 20
· - ·
0.25 . 35,000
-5.7lx10-6 1/h
Therefore, the component's remaining life, under the same creep The expected service life of the tube thinned down to 0.20 in would be
operating conditions, is (rounding down)
1
t = 5.71x10-6 {1 - [1+5.71x10-6x(4-l)x105 495]ldl-4 )}
'
01 nr
35,000 x · = 3800 h= 158 days
0.9 ::;; 50,554 h =5.8 yr
Because four years have elapsed, the remaining service life of the
10.17 Thinned Wall Remaining Life tube thinned down to 0.20 in is 1.8 years.
When creep causes wall thinning, such as is the case with overheated
boiler or furnace tubes, the remaining life may be calculated from11 10.18 Metallograph ic Life Assessment
Replicas can be examined by scanning electron microscopy to charac-
t =_!_11-l1 +K'(n-1)t ]lfCl-nl) terize the densi ty of creep voids, which is an indication of the degree of
nr K' r
creep damage. Models based on experimental results have been devel-
oped to correlate creep voids, determined by replication and metallog-
K':;; w;-wr
raphy, to remaining life. For example, in the case of 2"'.Cr-1Mo steel
W;Xt
welds, the number of cavities per square millimeter was correlated to
creep rate and time as follows,u
where tnr = remaining life of tube with wall thinning, h
K = r ate of wall thinning, 1/h
w, = intial wall thickness, in
wr = final wall thickness, at time t, in
t = time in service, h
where N = number of creep void cavities per mm2 , 1/mm2
t, time to rupture without wall thinning, h deldt = creep rate
n Norton law exponent, 4 to 8 for ferritic steel tubes
t =time, s
For example, if a tube in creep service had an initial wall thick-
In another model, also for weld heat-affected zones oflow-alloy steel,
ness Wi 0.25 in, and if the tube failed after four years of service (t =
the number of creep voids is correlated to the remaining creep lifell
35,000 h>. having thinned down to wr = 0.15 in,
0 25 0 15 t rem = texp 1 -1)

K':;; · - · =11.4x10-6 1 /h ( 1 _(1 -A)9
0.25x35,000
where trem = remaining life, h
With n = 4, thc time to rupture without wall thinning t, would lexp = expended life, h
be A = number fraction of ca\;tated boundaries
350 Chapter Ten
Chapter
11
For example, if after 50,000 h (5.7 yr) of creep service, 10 percent of
grain boundaries show creep void cavit ation, then A = 0.10, and the
remaining life is
1
t = 5o,ooo[ 1] = 31, 5oo h= 3.6 yr
rem 1-(1 -0.10)9
Overload
References
1. Thielsch Engineering, Cranston, RI.
2. Port, R. D., Herro, NALCO Guide to Boiler Failure Analysis, McGraw-Hill, New York.
3. ASTM E139, Starulard Thst Methods for Corulucting Creep, Creep-Rupture, arul Stress-
Rupture Thsts ofMetallic Materiais, ASTM Intemational, West Conshohocken, PA.
4. ASME Boiler and Pressure Vessel Code, Section li, Materiais, American Society of
5. API RP 579, Fit11Ess·for-Seruice, American Petroleum lnstitute, New York.
6. ASME Boiler and Pressure Vessel Code, Section Ill, Subsection NH, High Temperature
Compotu!nts, American Society ofMechanical Engineers, New York. 11.1 Overloads in Practice
7. Cohn, M. J., The influence of service temperature on creep rupture life, in ASl\.fE PVP
Conference, July 2004, San Diego. An overload, in the general sense, is any load (for ce, moment, displa ce-
8. API 530, Calculation of Heater-Thbe Thickness in Petroleum Refitu!ries, American
Petroleum lnstitute, Washington, DC. ment, rotation) beyond the design basis of the system, equipment, or
9. Cane, B. J., Remaining Life Estimation by Strain Assessment of Plant, Report component. In practice, the need to assess fitness-for-service of equip-
Rl/U2040 R81, Central Electricity Generating Board Research Laboratory, UK, 1981. ment subject to overloads appears in one of two ways:
10. Cane, B. J., Williams, J . A., Remaining life prediction ofhigh temperature materiais.
Int. Mater. Reu., Vol. 32, No. 5, 1987.
11. Viswanathan, R., Damage Mechanisms and Life Assessment of High Temperature • The actual overload. The equipment undergoes an accidental overload
Components, ASM Intemational, Metais Park, OH. that may or may not visibly deform it. For example, a pressure vessel
undergoes a pressure excursion well above its stamped maximum
allowable working pressure (MAWP).
• The postulated overload. An overload is postulated at the design
stage, as part of the plant or system safety analysis. This is the
"what if' overload. For example, an explosion may be postulated to
occur in a storage tank, and we need to know whether the tank will
remain leak-tight, or, ifit fails, how it will fail.
11.2 Overpressure Allowance

Accidental overpressure is probably the most common form of over-
load. E ach tank, vessel, piping system, and pipeline is designed to a
certain design pressure; unfortunately this design pressure takes dif-
ferent names, depending on the governing code: it is the MAWP for
ASME VIII pressure vessels, the design pressure for ASME B31 pres-
sure piping, the MAOP (maximum allowable operating pressure) for
pipelines, and the maximum fill height for storage tanks.
Systems are protected from overpressure by pressure-relieving
devices (typically gas safety or liquid reliefvalves, or ruptur e discs) or
by instrumentation and controls. This protection is to prevent exceeding
351
352 Chapter Eleven
Overload 353
the design pressure. Because the design codes include significant safety
factors, they also have allowances for exceeding the design pressure for
a limited time. For example, design codes permit the following pres-
sure excursions above the design pressure Pn.
• Power piping B31.1:
Pn + 15 percent for 10 percent of any 24 h
Pn · 20 percent for 1 percent of any 24 h
• Process piping B31.3:
Pn + 20 percent for 50 h, and 500 h per yr cumulative
Pn -t 33 percent for 10 h, or 100 h per yr cumulative
• Liquid pipelines B31.4:
Pn + 10 percent Figure 11.1 Rupture disk a>'sembly after incident.l
• Boilers:
MAWP + 6 percent
• Pressure vessel ASME VIII:
MAWP · 10 percent accumulation during reliefvalve discharge
MAWP + 121 percent accumulation during fire
• Nuclear pressure vessels ASME III:
Pn +- 10 percent during upset conditions
Pn +50 percent during emergency conditions
Pn + 100 percent during faulted conditions
11 .3 Overpressure beyond Allowance

Pressure can rise above the design overpressure allowance for several
reasons:
• The event that caused the overpressure was not foreseen at the design
stage, and therefore there were no measures in place to prevent or
relieve the pressure buildup.
• The overpressure event was foreseen, but the pressure relieving sys-
tem was not properly sized; for example, the relief de~-ice was too
small (not enough relieving flow area) or too large (reliefvalve became
unstable by open-close chatter). Figure 11.2 Postaccident condi-
tion of the vesse).l
• The overpressure event was foreseen and the relieving system was
sized correctly, but it was not well maintained. The relief valve may
have rusted to the point of causing so much friction that the valve
opened well above its original set pressure. investigation report noted that "The kettle was not equipped with safety
equipment, such as a quench or reactor dump system, to shut down the
A runaway chemical reaction occurred in a mixing vessel (a kettle) in process in case of a runaway reaction emergency. Also, pressure relief
a dye manufacturing plant. The reaction caused a rapid overpressure devices (rupture disks) were too small to safely vent the kettle in case
of the vessel, blowing a manway hatch off the top of the vessel and of runaway reaction."1 The pipe holding the rupture discs ruptured and
ejecting flammable vapors, causing an explosion and flash fires. The was ejected (Fig. 11.1) and a bolted top cover blew open (Fig. 11.2).
354 Chapter Eleven Overload 355
11.4 Key Consideratio ns for Overpressure consideration a design corrosion allowance. If the component has no
The key considerations for evaluating the integrity of tanks, vessels, and code pedigree its margin to failure is unknown.
piping subject to overpressure beyond their design basis are: • Does the component include nonweldedjoints? Are there bolted flanges
or manways, threaded joints, swaged or grooved joints, specialty fit-
• What are the component materiais? Are they ductile or brittle? What tings? Ex:perience teaches us that many overpressures have caused
are their mechanical properties? The integrity rules for ductile mate- failure not in the component base metal itself or its welds, but in its
riais will not apply to brittle materiais. The fitness-for-service assess- mechanical joints-manways or bolted vessel heads-which then act
ment will depend on the yield, strength, and toughness of the as pressure relief points (Figure 11.2).
materiais involved.
• What is the material condition of the component at the time of over- 11.5 Waterhammer Overload
pressure? Is it corroded or damaged? If the component is corroded or
damaged, then the rules for wall thinning, cracking, and geometric Overpressure can be due to waterhammer or, more generally, any forro
discontinuities will have to be applied as part of the overpressure of violent pressure transient. In fact there are typically three sources of
pressure transients:
analysis.
• What is the magnitude of overpressure, and what are the concurrent • Liquid transients (hammer)
loads? In particular what is the concurrent temperature? The mechan- • Liquid-vapor transients (bubble collapse)
ical properties of metallic materiais will depend on their temperature:
strength will decrease at high temperature; toughness will decrease • Liquid-gas transients (slug, stratified, bubble flows)
at low temperature. The integrity of nonmetals, such as seals, gaskets, In a single-phase liquid system the transient is a waterhammer. It
0-rings, and packing, will be significantly degraded when hot or very is caused by a sudden change in flow velocity causing a corresponding
cold. change in pressure. When liquid flow is stopped "slowly," for example,
• How fast is the overpressure? Is it a quasistatic steady pressure rise, when closing a manually operated multiturn valve, the velocity changes
ora deflagration ora detonation? This will affect the behavior ofthe from the normal operating flow velocity to zero at flow stoppage. This
equipment. Separate sections in this chapter cover integrity of com- happens slowly and the pressure rise is quite small.2
ponents subject to deflagration and detonation.
dP = p(ôu)2
• What is the phase of the contained fluid? Is it a gas, a vapor, a flash-
ing liquid, or a nonflashing liquid? The phase of the fluid contained 2g
will have no effect on the onset of failure; in other words, identical
components containing liquid, steam, or gas will rupture at the where dP = pressure change at disturbance, psi
same limit pressure. But the failure mode, what happens once the p = fluid weight density, lb/in3
fracture starts, will vary greatly. The worst failures tend to be with ~v = change of fluid velocity at disturbance, in!s
flashing liquids: for example, a rupture in a pipe carrying pumped g = gravity = 386 in/s2
condensate at, say 400 psi and 300°F, will cause the condensate to
For example, if a valve is closed slowly on water at ambient tempera-
violently flash to steam, which will accelerate the opening ofthe ini-
ture (p = 62.4 lb/ft3) flowing at 10 ft/s, the pressure rise in bulk flow is
tial crack.
• Is the component fabricated to a code or standard? Are the initial dP=p(ôu)2 = 62.4x102 1 =0.7psi
qualities of machining, welding, and inspection adequate? Systems, 2g 2 X 32.2 144
equipment, and components designed and fabricated to a code have
an initial safety margin and inherent quality. For example, a pressure But if the flow stoppage is fast, faster than 2Lia, where L is the dis-
vessel designed and fabricated to the ASME VIII Di v. 1 code in 2004 tance from the source of pressure (a pump ora large header, Fig. 11.3)
has an inherent safety margin of 3.5 against rupture, taking into to the closing valve, and a is the sonic velocity in the liquid, then the
Figure 11.3 Distance L from pre,;sure source to valve.
pressure rise could be large, causing a loud hammerlike bang and jerk-
ing the line
dP= pa(w}
g
where dP - pressure change due to instantaneous valve closing, psi

p = fluid weight density, lb/in3
ôu - change of fluid velocity at disturbance, in/s
g = gravity 386 inls2 Figure 11.4 Rupture of cast iron steam valve at neck.
a = sonic velocity in liquid, inls
In the above example, halting a 10 ftJs flow, if the flow stoppage is
faster than 2L/a, the pressure rise would be liquid with trapped air. Rather than hammer overload, what takes place
here is vibration particularly in continuous bubbly flow (the turned-on
dP= pa(óu) = 62.4x4860x10 1 = 654 psi garden hose syndrome).
g 32.2 144
11.6 Bolted Joint Failure
The second type of transient occurs in a two-phase liquid-vapor flow
regime, when the liquid could evaporate and, vice versa, the vapor could Following an aborted startup at a polymer plant, hot molten plastic
condense to liquid. The classic example is a bubble collapse hammer in was diverted to a catch tank. Unknown to the operators, the material
steam-condensate lines. The steam if surrounded by cool water will con- started to foam and expand inside the vessel, clogging the vent pipe
dense to water, leaving a vacuum in the space it occupied. This "bubble" and pressure gauges mounted on the vent. Based on prior experience
is violently filled with the surrounding condensate that bits the pipe wall with the system, and possibly an erroneous reading from the clogged
causing a hammer. If the material is brittle, for example, a cast iron pressure gauge. the tank was readied for opening and cleanup. With
valve body, it may fracture from the dynamic pressure rise, discharging half of the 44 bolts removed from the tank head, the cover blew off vio-
steam. This is what happened in the case illustrated in Fig. 11.4 where lently, spewing hot plastic.3
the steam valve ruptured at its neck, and the pipe that was on the left- The integrity and leak-tightness of a bolted flange joint, such as a
hand side of the valve was propelled into the air causing serious damage. manway, is addressed in ASME VIII Div.l Appendix 2. To achieve a
Where the pipe once stood, only the concrete pedestal remained. stable leak-tight joint, the bolt preload W,. must be larger than a leak
The third type of transient occurs in a line filled with liquid and gas; tightness minimum preload Wm 1 plus a gasket compression minimum
here one phase does not transform into the other. A classic example is preload wm2·
H is the force exerted on the head of diameter G, and is equal to
H= nG2 p
4
where P = applied pressure, psi and G = diameter of bearing line, in.

For example, a vessel head flange has a diameter G = 4 ft and expe-
riences an accidental overpressure to 600 psi; the hydrostatic force is
H = 1t X (
4 X 12)2 X 600 L 1, 086, 000 c 1 X 106 lb
4
A 4-ft vessel is not unusually large, and 600 psi is not an extraordi-
narily high pressure, yet the pressure force on the bolts is huge, over 1
million pounds, enough to send a vessel head flying once the bolts are
incapable of holding the pressure. In service, in addition to the force H
Figure 11.5 Vent pipe clogged with polymer.3 needed to counteract the internai pressure, a force Hp is needed to
compress the gasket; this compression force is
H p = (2b) (nG) m P
where b = effective gasket width, in, and m = experimental proportion-

ality constant.
In the example, if the gasket is spiral-wound fiber-filled stainless
steel, with an effective gasket width of 0.5 in, with m = 3.0, then the
force needed to compress the gasket is
Hp = (2 X 0.50) X (rt X 48) 3.0 X 600 = 271,000 lb.
The total preload needed to hold the pressure is
Wml = 1,086,000 + 271,000 = 1,357,000 lb

In addition the preload must also compress the gasket material to a min-
imum stress y that depends on the type of gasket, and can be obtained
from the gasket manufacturer,
Figure 11.6 Catch tank without head cover.3 W m2 = (nG)by
In our example, if the seating stress for the gasket is given by the
manufacturer as y = 10 ksi, then the preload to seat the gasket is
Wm2 = ('IT X 48) X 0.5 X 10,000 = 754,000 lb

where wm 1 bolt preload to sustain the design pressure, Ih
=
H = bolt preload to resist the hydrostatic load in the pres- The required bolt preload is therefore
surized pipe, lb
Hp = experience based preload to ensure leak-tight joint, lb Wm = max (1,357,000 lb; 754,000 lb) = 1,357,000 lb- 1.4 million lb
11.7 The Bullet Pig

If a 20-in diameter drying or cleaning pig is propelled through a pipeline
at 400 psi, the force propelling the pig is
D2 202
F=n-P=n-x400-126,000 lb
4 4
If the pig is received at an open trap, an applied force of this magni-
tude, multiplied by a dynamic impact load factor, could cause quite
large damage, as illustrated in Figs. 11.7 and 11.8 where a backhoe was
placed at the open end of a pipeline trap to receive a pig.
11.8 Detonations and Deflagrations

The need to analyze the integrity of a tank, vessel, or pipe subject to
explosive pressures may arise during design or safety analysis of pos-
Figure 11.8 Backhoe meant to stop the pig.
tulated accidents, or when investigating an accident. First, let's clarify
the terms explosion, deflagration, and detonation, by simply referring
to the National Fire Protection Association standard NFPA 68:
the speed of sound in a stochiometric mixture of air and hydrogen is
Explosion. "The bursting or rupturing of an enclosure or a container approximately 1400 ft/s (1000 mph). Note that the propagation of a
dueto the development ofinternal pressure from a deflagration." We much slower combustion front, less than 100 mph, is referred to as a
could add at the end of this definition "ora detonation." "fireball."
Deflagration. "Propagation of a combustion zone at a velocity that is Detonation. "Propagation of a combustion zone at a velocity that is
less than the speed ofsound in the unreacted medium." For example, greater than the speed of sound in the unreacted medium." For
example, the measured speed of a hydrogen-air detonation front in a
2-in pipe is near 7000 ft/s, clearly above the sonic velocity of 1400 ftls.
As deflagrations travei down pipes, within 10 to 20ft they tend to

transition to detonations, particularly if
• The deflagration front is confined or mixed

• The explosion is initiated by a strong energy source (such as a light-
ning strike)
11.9 Explosion Pressures

A detonation starts with a sudden pressure impulse (the "ZND" spike)
at nearly 30 times the initial pressure P 0 • This spike is of very short
duration, on the order of 10-4 to 10-5 seconds for hydrogen-air mix-
tures. The pressure drops to a CJ pressure around 15P0 , and then
reaches a steady residual pressure around 8P0 , until vented.
The peak pressure of a detonation in a pipe is related to the flame
Figure 11.7 Open trap. front velocity
sections such as passages through valves and reducers, and at dead

ppeak = (1 + y-1)(!:)2- y-1 ends, such as blind flanges.
P0 y+1 a y+1
where Ppeak = detonation peak absolute pressure, psia The peak pressures caused by deflagration are lower than those
Po = initial absolute pressure inside the pipe, psia caused by detonation, in the order of eight times the initial absolute
pressure.
'Y=e/cu
Cp = heat capacity at constant pressure
cv = heat capacity at constant volume 11.1 O Explosion Damage
v = flame front velocity, ft:Js
a = sonic velocity, ft:Js The extent of damage from explosions in tanks, vessels, and pipes
dependa on the dynamic response (regime) ofthe component, which in
For example, if the detonation in a pipe containing a 30% hydrogen-air turn depends on two key parameters:
mixture has a flame front velocity v =7000 ft:Js, then
• The magnitude of the pressure spike
2
p peak =(1 + 1.4 -1)(7000) - 1.4-1 c 29 • The duration of the pressure pulse Tputse compared to the natural
P0 1.4+1 1400 1.4+1 period of the component Tcomponent
The detonation peak pressure is, in this case, 29 times the initial There are three regimes of dynamic response:4
absolute pressure. For example, if the hydrogen-air mixture was ini-
tially at atmospheric pressure, the peak detonation pressure would be • The quasistatic regime
29 X 14.7 psia = 426.3 psia = 411.6 psig. Unfortunately, in practice, • The impulsive regime
things are not so simple:
• The dynamic regime
• To predict the peak pressure using the formula above, we need to
know the flame front velocity v, which we do not have at the onset. The quasistatic regime occurs if we have a relatively stiff component
(high natural frequency, low natural period, e.g., a small diameter, stiff
• Detailed analytical predictions of explosion pressures indicate peak pipe) subject to an explosion of long duration; it occurs if4
pressures double the above predictions at orifices, at narrowing cross
Tpulse > 40 Tcomponent·
30P0
ZND The structural effect of the explosion will be equivalent to the peak
p pressure statically applied to the component, multiplied by a dynamic
amplification factor. The dynamic amplification factor would be 2.0 for
a step pressure, and 1.5 for a triangular pressure.õ
The impulsive regime occurs in the case of a relatively flexible com-
15P0 - - - - - CJ ponent (e.g., a large storage tank) subject to an explosion ofvery short
duration, such as a detonation; it occurs if
Tpulse < 0.4 Tcomponent

8P0 -- --- ------~--------------
The component has barely started to deform, let alone reach its fully
stretched condition, before the peak pressure has disappeared. The dam-
Pol-----' aging effect is nowhere near the quasistatic regime effect. The damage
no longer dependa on the peak pressure, but instead it dependa on the
impulse ofthe explosion, the area under the pressure-versus-time curve,
Figure 11.9 General trend of hydrogen detonation pressure. such as shown in Fig. 11.9. The dynamic regime is a transition regime
between quasistatic and impulsive, where the effect is govemed both by 11.13 Explosive Rupture and Fragmentation
the peak pressure as in the quasistatic regime and the impulse energy as
We first look at the pressure that could cause ductile rupture of a cylin-
in the impulsive regime; it dominates if
drical shell of diameter D and thickness t. The burst pressure can
40 T component > Tputse > 0.4 T component · be calculated using the Cooper formula8
11.11 Exampi~Deflagration in Pipe

What is the effect of a 1-s hydrogen deflagration, with a peak pressure
of 200 psi, inside a 1.5-in schedule 80 steel pipe? We must first estab- where P burst = burst pressure, psi
lish the explosion regime. The pipe's radial breathing mode natural Su = ultimate tensile strength, psi
frequency n = strain hardening coefficient of the material
This formula can also be written as the Langer burst pressure formula9
f=-1- ~= 1 28x10s = 36 000 Hz
21tRV~ 2x1tx0.9 7.53xl04 x(1 - 0.32 ) '
Pburst=2.31x O.577
n
x--n
txSu
T romponent = 7
= 2.8 X 10-5 S
An altemate estimate is obtained using the Svensson burst pressure
formula 10
where f = natural frequency, Hz
E
R = radius, in
= Young's modulus, psi -
p burst -
X
Su n + 0.227 n
(e
0.25 X - )n X2t-
D
Pm = mass density, lb.s2/in4
u = Poisson ratio We analyze next the conditions under which an explosion can cause frag-
mentation and shrapnel to tear out of the metal wall of a vessel, tank, or
Because 1 s > 40 X 2.8 10- 5 s, the pipe response is in the quasistatic pipe.5,u Consider that the explosive pressure P accelerates radially a
regime. At 1-s duration the deflagration is sufficiently long to fully small volume ofthe component wall mass (Fig. 11.10).29
stretch the pipe wall. The pressure felt by the pipe wall will be as high
as 400 psi, twice the actual pressure. In this case, the pressure can f'..m x r = Px!::J.Lxr x de
readily be sustained by a 1.5-in schedule 80 steel pipe.
A precise analysis of tanks, vessels, and piping subject to explosions
can be achieved by applying the time-history profile of the explosion
pressure (P-versus-t) to an elastic-plastic finite element model of the
component.
11.12 Material Strength at High Strain Rates rI '

I I
.:4Z>
'
At high strain rates of the magnitude encountered in explosions, the I

I
yield and ultimate strength of metais increases.2.6,7 A compendium of
I '
mechanical properties-versus-strain rate has been recently compiled. I
I
0'6i,'
'
For mild steel ata train rate of !1EIM = 0.2/(2 X 10- 4) = 1 X 103 1/s, at I
1\,
I
I ' Figure 11.10 Volume element of
room temperature, the compendium indicates an increase in ultimate I '
component wall.
I''
strength from 55 ksi at static loading up to 80 ksi at 103 1/s, a 1.45 ,,
I,
increase. N
/lm = p X t X ôL X r X de
p X t X ôL X r X de X r = p X jL X r X de
p PxT
r=-- r=--
pxt pxt
Fragmentation takes place when the energy imparted by the explo-

sion exceeds the strain energy that the metal can absorb as it is
stretched radially from r0 to r0 (1 + E)
Figure 11.11 Ductile rupture from shock wave.
1 .,
2x~xr- >Pr xM.xr0 XE
Finite element analysis is the only viable method for shapes other
2 than cylinders or spheres away from structural discontinuities, and
1
-(pxtxôLxrxde) (PxT) S xt
- - >-'-xõLx rxd9xr0 XE where the component is subject to cyclic pressures. Plastic design rules
2 pxt r0 are currently in development for the analysis of pressure vessels subject
to explosive pressures (deflagrations or detonations). A first tentative
design rule to prevent ductile fracture is to limit the plastic strains in
the wall of impulsively loaded steel vessels to the following
where P = explosive pressure. psi Em :S 0.2%
t - wall thickness, in
T =time interval when metal is exposed to pressure wave, s Eb :s 2% in base metal and 1% at welds
S 1 flo\\' stress of the material, psi
Ep :s 5% in base metal and 2.57t at welds
Pm =material mass density, lb-s2 -in4
E - strain at failure where Em = maximum in-plane plastic strain averaged through the
ôL - length of small metal section, in thickness of the vessel
de = angle at origin of small metal section, rad Eb = maximum in-plane plastic strain linearized through the
M =face area of small metal section, in2 =óL r de thickness of the vessel
r = radius from centerline of small metal section, in Ep = maximum peak equivalent plastic strain at any point in
Pr = pressure causing a flow stress in the metal, psi the vessel
This method should be only applied when the hoop stress due to the For vessels subjected to multiple impulsive loading events, the in-
explosion exceeds the flow stress of the material. The fragmentation plane plastic strain components must be accumulated over successive
formula correlates well with detonation damage, in the impulsive Ioading events. In addition, to the design strain limits above, the follow-
regime, where Tis very small compared to the vessel period. Under ing cautions and rules apply to the analysis of detonations in vessels,
slower, dynamic, or quasistatic pressures, the rupture of ductile mate- tanks, and pipes:
riais will be by longitudinal split (fish mouth), and if the explosive
energy is sufficiently large, the rupture ends will slant at 45 degrees, • Tbe designer shall consider the need to reduce these strain limits for
along lines of maximum shear stress (Fig. 11.11). areas of high biaxial or triaxial tension.
Tbe closed form solutions described so far work well for simple shapes: • A fracture mechanics fatigue evaluation should be conducted, for
cylinders or spheres, away from discontinuities. For ali other shapes, the example, following the methods and requirements of ASME VIII
integrity of tanks. vessels, and pipes subject to internai explosions Division 3, Article KD-4. This is particularly necessary for low-
should be established by detailed finite element analysis or testing. temperature applications. thick-wall components, steels with low
toughness, high-strength steels, or steels not listed in ASME II, where li = midspan deflection, in
and where the component contains preexisting cracks. lJ = diamete~ in
L =span length, in
Other plastic design methods for large pressures are provided in
p, = mass density of pipe Ob·slin'1)
ASME III, Div.1 Appendix F, andASME VIIT, Div.3. Ifwe are interested
A = metal cross-sectional area, in2
in predicting the actual burst pressure by finite element analysis, rather
E = Young's modulus, psi
than applying design rules with safety margins, then the computed
I = moment of inertia of cross section, in4
elastic strains should be compared to ductile fracture strains
E = strain
n n
E
' .bunt-cyl
=-
J3 Ehoop,burst-cyl = 2 11.15 Natural Phenomena Hazards
Earthquakes, high winds, and landslides are sources of overload on
where E..bu....t-cyl = effective strain at burst of cylinder (for example, tanks, vessels, piping systems, and pipelines. Building codes prescribe
von Mises strain at burst) when and how to design for these hazards. For many years, building
Ehoop.bun~t-cyl = hoop strain at burst of cylinder
codes focused on buildings, dams, and bridges. In the last decade they
n strain at necking of a tensile test specimen have also developed roles for critica) equipment and systems.12
For a sphere, the relationship is The effect of earthquake overloads is twofold:
• An inertial effect that causes equipment and systems to shake, sway,
n
Ehoop,bun;t.,.ph =J vibrate, and possibly rock, shift, overturn, leak, or break. In Fig.
11.12, a pump skid slid off its vibration isolators due to large lateral
earthquake accelerations. In Fig. 11.13, earthquake shaking caused
where E.,bW"!'t-sph - effective strain at burst of sphere (for example, von one pipeline to uplift and the other to slide, ending up in the position
Mises strain at burst), and Ehoop.bun;t-sph = hoop strain at burst of sphere.
11.14 Effect of Externai Explosions

The overload from an externai explosion on tanks, vessels, and pipes
consists of an incident blast pressure wave pushing the component
away from the origin of the blast, and a reflection towards the origin
of the blast if the component is near a rigid wall, and a reversing pres-
sure on the back si de of large tanks and vessels, towards the origin of
the blast. Damage to simple configurations such as pipe or pipeline
spans can be calculated. In the impulsive regime, the elastic-plastic
deflection of a pipe span under a side-on blast pressure is5
iDL2
li--.=======
- ~24.576 X 2 X pAEJ
The corresponding maximum strain in the component wall is
Figure 11.12 Seismic failure ofvibration-isolating springs.

Figure 11.13 EtTect of seismic shaking on pipeline.
Figure 11.15 Ground and peer settle below pipeline.
bination of analysis and testing, or the use of earthquake experience-

based rules. The more difficult task is not to seismically design the sta-
tic equipment tanks, vessels. pipes) but to provide for operability of
criticai active equipment (pumps, compressors, fans) and their power
supply and instrumentation and controls.
Figure 11.14 Aftennath ofearthquake. The fitness-for-service oftanks, vessels, and piping subject to natural
phenomena hazards is accomplished primarily by analysis, focusing
particularly on bracing and anchorage. 13 There are many codes, stan-
shown. There was no leak, which is a tribute to the material's duc- dards, and publications that address seismic design, retrofit, and
ti1ity and the weld quality. mtegrity of storage tanks, vessels, p1ping, and pipelines. 14-24 The attrib-
• An anchor motion effect that causes large differentíal movements utes generally considered in fitness-for-service of equipment subject to
between the equipment support points. These effects are often caused natural hazards are:
by permanent shifts of the supporting structure or ground. Figure
11.14 shows a seismic-induced ground rupture, the kind that would • Storage tanks. Bending stress on walls, uplift, and anchorage, free
impart a very large tensile and bending load on buried pipelines. board, and sloshing of contained liquids
Figure 11.15 shows how seismic-induced ground settlement shifted the • Pressure vessels. Support legs, attachments between vessel and sup-
valve support down. After the earthquake, wooden shims were added ports, anchorage
to support the sagging valve.
• Towers and columns. Same as pressure vessels plus shell bending
In earthquake-prone areas, storage tanks, vessels, piping, and • Piping systems. Pipe stresses, supports and anchors, equipment noz-
pipelines in criticai service can be seismically designed either by a com- zles, loads at mechanical joints, and, where required, operability of
active mechanical equipment and components (pumps, compressors, carried on until collapse (hinge, excessively large defonnation) is
valve operators, etc.) achieved at a load Lrouaptw'· The limit on the collapse load is
• Buried pipelines. Seismic wave passage, soil settlement or failure,
L < 0.9 Lconapse
settlement of end points
where L = applied load, lb or in·lb, and Lcollap.ie = collapse load, lb
Seismic and wind design rules can be followed for the fitness-for-service
or in-lb.
assessment of existing equipment, but the fitness-for-service assessment
must account for the initail material condition: wall thinning, cracking, or Plastic collapse analysis. A plastic collapse analysis relies on a plas-
damage. tic model, with strain hardening. The load is applied to the model
and the limit load is defined as the load at which the deformation or
11 .16 Fitness-for-Service by Plastic strain reaches twice the deformation at the onset of yield
or Collapse Analysis
<1>2 = 2 tan - 1<1>1
ASME 111, Appendix F provides u seful guidance for the analysis and
assessment of the integrity of pressure equipment under large loads, where <1>2 = strain or defonnation at limit load, and <1>1 = strain or
beyond the elastic limit. A summar y of the tive Appendix F assessment deformation at onset of yielding
techniques and criteria follows.
Plastic instability analysis. Th e plastic model is progressively
Elastic analysis. The stresses are calculated elastically, even though deformed, for ming multiple hinges, to the point where plastic insta-
the actual stress due to overload is plastic. The limits on the elastically bility occur s:
calculated stresses are:
L < O. 7 Lin8tability
P,., < 0.7 Su
where L = applied load, lb or in-lb, and Lm.stability = instability load,
Pm + Pb < 1.05 SL lb or in·lb.
T < 0.42 Su
11.17 Bending Failure
where Pm = primary membrane stress, psi
P = primary bending stress, psi We know, from bending small copper tubes by hand, and bending large
Sr = ultimate strength, psi steel pipes in the shop or in the field, or from accidental bends during
T =shear stress, psi handling or installation, that ductile materiais rarely rupture when
bent. They do buckle, as illustrated in Fig. 11.16, wher e a 90° tee (a
Plastic· analysis. This is a plastic analysis where the material is nozzle) was subject to large in-plane bending towards the right, causing
modeled with its actual stress-strain curve. The limits on plastically the compressive side of the nozzle-header, at right, to buckle inward.25
calculated stresses are The buckle itself could also be the source of a crack, as explained in
Chap. 8. But sometimes bending could cause a rupture. There are three
S < 0.9 Su
conditions under which this is possible:
Pm < 0.7 Su
• Bending a ver y thick wall component could cause the outer fiber to
T < 0.42 Su reach the ultimate strain. However, bending very thick parts occurs
very rarely.
where S = primary stress intensity, psi. • Bending a badly corroded wall.
Limit collapse analysis. The material is modeled as elastic-perfectly • Bending accompanied by large tension. This happens in buried pipes
plastic with an elastic stress-strain followed by a flat horizontal subject to landslides or large soil settlements. The moving ground
stress-f;train line beyond yield (no strain hardening). The analysis is places large tensile loads on the pipe, which add to the bending
2. Antak:i, G. A., Piping and Pipeline Engineering, Dek.ker, New York.

3. Thermal Decomposition lncident, U.S. Chemical Safety and Hazard Investigation
Board, Washington, DC Investigation Report, Report No. 2001-03-I-GA, 2001.
4. Baker, W. E., Cox, P. A., Westine, P. S., Kulesz, J. J., Strehlow, R. A., Explosion
Hazards and Evaluation, Fundamental Studies in Engineering, Vol. 5, Elsevier,
New York, 1983.
5. Baker, W. E., The elastic-plastic response of thin spherical shells to intemal blast
loading, J. Applied Mechanics, March 1960.
6. Rinehart, J. S. and P earson, J., Behavior of Metais under Impulsiue Loads, American
Society of Metals, 1954.
7. Bitner, J. L. and Hampton, E. J .. Stress or Strain Criteria for Combined Static and
Dynamic Loading, Pressure Vessel Research Council.
8. Cooper, W. E .. The Significance of the Tensile Test to Pressure Vessel Design,
Welding Research Supplement, January 1957.
9. Langer, B. F., WRC Bulletin 95. PVRC lnterpretiue Report ofPressure Vessel Research,
Section 1-Design Considerations, by B.F. Langer, Pressure Vessel Research Council,
April1964.
10. Royer, C. P., Rolfe, S. T., Effect ofstrain hardening expqnent and strain concentrations
on the bursting behavior of pressure vessels, Transactions of the ASME, Joumal of
Engineering Materiais and Technology, paper No. 74-Mat-1 presented atASME PVP
Figure 11.16 In-plane bending collapse of90° tee.25 Conference, 1974.
11. Price, J. W.. An acetylene gas cylinder explosion, Trans. ASME, Vol. 120, February
1998.
12. ASCE 7, Minimum Design Loads for Buildings and Other Structures, American
Society of Civil Engineers, 2002, Reston, VA.
13. Bednar, H. H., Pressure Vessel Design Handbook, Krieger, Melboume, FL.
14. API 650. Welded Steel Tanks for Oi/ Storage, Appenclix E, Seismic Designo{ Storage
Tanks, American Petroleum Institute, Washington, DC.
15. AWWA - D100, Standard for Welded Steel Storage Tanks, American Water Works
Association, Denver, CO.
16. Seismic Design and. Eualuation Guidelines for the Department o{ Energy High-Leuel
Waste Storage Thnks and Appurtenances. BNL-52361, 1995, Brookhaven National
Laboratory, New York.
17. ASME Boiler and Pre.ssure Vessels Code, Section VIII, Pressure ~ssels, American
Society of Mechanical Engineers, New York.
18. ASME Boiler and Pressure Vessel Code, Section III, Division 1, Nuclear Components,
Subsections NB/NC/ND-3600, American Society of Mechanical Engineers, New York.
19. Seismic Design and Retrofit of Piping Systems, American Lifelines Alliance, 2002,
Washington, DC.
20. ASME B31 Code. Pressure Piping, American Society of Mechanical Engineers, New
York.
21. NFPA-13, Sprinkler Systems, National Fire Protection Association, Quincy, MA.
22. Guideline for the Design of Buried Steel Pipe, American Lifelines Alliance, 2001,
Washington, DC, www.americanlifelinesalliance.org
23. Guidelines for the Seismic Design of Oi/ and Gas Pipeline Systems, American Society
of Civil Engineers, 1984. Reston, VA.
Figure 11.17 Girth weld crack from soil movement. 24. ASCE 4 Seismic Analysis of Safety-Related Nuclear Structures, American Society of
Civil Engineers, 1984, Reston, VA.
25. WRC Bulletin 219, Experimentallnuestigation ofLimit Loads ofNozzles in Cylindrical
stresses, eventually causing a circumferential failure, as illustrated Shells, September 1976.
in Fig. 11.17, particularly at girth welds that are not fully penetrated 26. Biggs, J. M., lntroduction to Structural Dynamics, McGraw-Hill, New York, 1964
or that may have fabrication flaws. Note in the figure the buckle on 27. WRC 347, (1) Welded Tee Connections of Pipes E:cposed to Slowly Increasing Internai
Pressure, by J. Schroeder and (21 Flawed Pipes and. Branch Connections Exposed to
the bottom compressive side. Pressure Pulses and Shock Waues, by J . Schroeder. Pressure Vessel Research Council,
September 1989.
28. Kiefner & Associates, Worthington, OH.
References 29. C. McKul, personal communication.
1. Chemical Manufacturing Incident, U .S. Chemical Safety and Hazard
Investigation Board, Washington, DC, Report No. 1996-06-I-NJ, 1996.
Chapter
12
Failure Analysis
12.1 Failure Mode and Effects

This chapter is dedicated to understanding the mode and effects of
failures of tanks, vessels, piping, and pipelines. Failure can-in the
broadest sense---occur in one of three ways:
• Loss of stability. The component becomes unstable, sags, buckles,

falls, rocks, or overturns. For example, over time, spring supports
lose their stiffness and a pipeline sags, or the ground settles under
a storage tank.
• Loss ofpressure boundary. The equipment ora component leaks or
breaks. This failure can be progressive, for example, due to corro-
sion, or sudden, for example, as a result of a rapid overpressure or
explosion.
• Loss of operability. The component no longer performs its function, or
conveys and controls flow as intended. For example, heat exchanger
or boiler tubes become plugged with deposits, or isolation valves are
no longer leaktight.
The effects of these failure modes may be acceptable, and accounted
for in design, or they may be unacceptable. A leak may be acceptable
in low-pressure steam but not in high-pressure steam. A leak in a
water storage tank may be acceptable, but not a leak in a tank storing
chlorine.
Where the effects of failure are significant in terms of safety and
cost, several precautions apply:
• Materiais. Choice of corrosion-resisting alloys, lining and coating;
strict material control and traceability
an
378 Chapter Twelve Failure Analysis 379
• Design. Redundant and high-quality instrumentation and controls, • Was the toughness (CVN, KIC, etc.) unusually low?
alarms, use of double-wall tanks and pipes, use of ali welded construc- 2. Design
tion, formal safety analysis of facility and development of normal and 2.1. Basic System Design
upset design conditions, design to prevent or mitigate extreme loads • Did the basic process work as expected?
such as postulated explosions • Did an unexpected chemical or physical reaction take place?
• Construction. Personnel and process qualification, augmented quality • Was the safety logic sufficient and operational?
control and third-party audits, 100 percent rather than spot radio- • Was the throughput within limits?
graphy, pressure test and sensitive leak test rather than in-service • Were the instruments and controls suffi.cient and operational?
leak test, as-built verification of construction against design • Were instruments and controls properly calibrated?
• Was overpressure protection adequate and did it function?
• Operation. Operator training, formal procedures, operationallimits
• Did alarms work as planned? Were they sufficient?
and alarms, emergency response procedures, up-to-date system 2.2. Detailed Integrity Design
drawings and procedures
• Did the system experience an overpressure beyond its
• Maintenance. Formal predictive inspection program (structural rating?
integrity prograrn), analysis of results and fitness-for-service within • Were layout and weight support adequate?
prescribed schedules, immediate reporting of nonconformances and • Was there sufficient flexibility in hot or very cold systems?
follow-up, formal repair program, strict update of design and opera- • Was thermal shock or fatigue at play?
tions documents, control and multidiscipline review of modifications, • Was there an unusually large levei of vibration?
lessons learned program, preparation and maintenance of system • Did thermohydraulic transients occur?
health reports • Was there high-temperature creep?
• Was there an unusual natural hazard (soil settlement, high
12.2 Root Cause Failure Analysis wind, etc.)?
3. Construction
There are many techniques for the analysis offailure modes and effects; 3.1. Fabrication and Erection
but for tanks, vessels, and piping systems, we should refer back to Chap. • Were forrning and machining quality acceptable?
1, which lists the five key areas of conduct of operations: materiais, • Was welding quality within code?
design, construction, operation, and maintenance. These same five • Were abnormal loads introduced during handling, aligning,
areas provide the basis to generate lines of inquiries to investigate the and erection?
root cause of failure of tanks, vessels, piping, and pipelines. • Was mechanical joining per vendor requirements and codes?
1. Materiais • Were flanged joints properly assembled and bolted?
1.1. Material Selection • Was construction per design; have deviations been accepted
• Was the material compatible with the process chemistry? by engineering?
• Was the material compatible with the operating pressure, 3.2. Inspection and Testing
temperature, and flow rate? • Was NDE per code? Was it the right technique, by the right
• Is there evidence of corrosion not accounted for in design? people, to the right criteria?
• Was the corrosion rate larger than the design corrosion • Was leak testing conducted; was it adequate?
allowance? 4. Operation
1.2. Material Quality 4.1. Instrumentation and Controls, Procedures, and Training
• Was the material chemical composition per spec.? • Was the product stream (on the inside) same as basis of
• Was the microstructure as expected? material selection in 1.1?
• Was the strength (yield, ultimate, elongation at rupture) per • Was the environment (on the outside) same as basis for
spec. mínima? material selection in 1.1?
• Are the component dimensions and finish in accordance with • Are the operating parameters (pressure, temperature, flow)
the specification? same as designed for in 2.2?
380 ChapterTwelve Fallure Analysls 381
• Were the operating procedures adequate; were they followed Design drawings and calculations
by operators? Fabrication, NDE, and pressure test records
• Were operators trained and familiar with the system function Operationallogs and operator interview
and emergency response? Valve alignments, flows. instrument readings
• Was there an overload (demand in service) that exceeded the Markup system condition at time of accident on P&ID
design basis loading: weight, pressure, temperature. vibration. Maintenance records and maintenance interview
transient, externai impact. and so on? • Visual inspection of scene
4.2. Emergency Response Direct visual, magnifying glass. and light
• Did the operators follow the emergency response procedures? Tape measure
• Were the emergency response procedures adequate? What failed and what did not
• Were operators knowledgeable of the system condition and Other defects in the vicinity
its risks? Ali credible failure modes; do not eliminate at this stage
• Was there unsafe behavior? Failure effects
• Did production come ahead of safety? Photos and sketches
5. Maintenance Quality of fabrication (welds, joints, etc.)
5.1. Risk-Based Inspection and FFS Quality of maintenance and component condition
• Was this similar to a previous failure or degradation that was Replication of metal surface
overlooked? Samples for laboratory testing, label samples
• Was this a failure due to poor prior repair? Collection of corrosion samples and other deposits
• Were instruments and controls maintained, tested, and cali- Collection of process fluid sample for corrosion analysis
brated regularly? Cleaning the surface with dry air
• Was performance and failure history missed or inadequate? Brushing off or ultrasonic cleaning adherent deposits
• Did in-service inspection (1\TJ)E) take place? Was it adequate? Solvent cleaning (HCI on carbon steel, sulphate and hydrofluoric acid
• Were lhe results of inspections analyzed for fitness-for-service on SS)
(FFS>?
• Were FFS recommendations implemented? • Nondestructive examination
5.2. :\fanagement of Change Liquid penetrant testing. magnetic particles testing
• Was there inadequate control and management of change? Radiography, ultrasonic testing. eddy current testing
• Are drawings and operating procedures kept up to date to Wall thickness
reflect changes? Hardness
• Was this a failure due to prior repair or change? • Fractographic examination
• Does the company have a good management of change Low-power magnification, stereomicroscope 5X to 50X
'MOC) program (Chap. 1)? Optical microscope 2000 ..
• Was the prior repair or change conducted according to the Transmission electron microscope of replica 50,000 x
MOC program? Scanning electron microscope 50,000X
Failure mode: brittle, ductile, microvoid coalescence, fatigue stri-
12.3 Failure Analysis Tools
ations, plasticity
• Chemical analysis of process fluid
The keys to a successful investigation are (a) keep an open mind ear- General chemistry
ly on, not discarding a possible cause of failure until proof is obtained, Trace impurities
and (b) follow a structured process. The following steps apply to the Comparison \vith past history
investigation of failures in tanks, vessels, and piping.
• Metallurgical tests
• Background information Energy dissipative X-rays for chemical composition of metal
Material certificates Chemical analysis of surface deposits, scale. and oxides
382 Cha pterTwelve Failure Analysis 383
Inspection for inclusions on polished surface

Metallographic examination of polished and etched specimen
Intergranular or transgranular cracking
Microhardness
Strength tests: yield, ultimate, elongation at rupture
Toughness tests: CVN, KJC, CTOD, Jc, DWTI'
• Stress analysis
Estimate forces, pressures, energies that caused observed damage
Estimate stresses or strains in component and explain behavior
Hand calculations or finite element analysis
Elastic or plastic analysis
Creep analysis Figure 12.1 Pinhole leak.
Leak-before-break and fracture mechanics analysis
• Corrosion testing in laboratoryi-30
Immersion
Coupon
Weight loss
Visual
Anodic or cyclic polarization
Electrical resistance probe
Galvanic probe
Linear polarization resistance probe
Corrosion potential
Intergranular
Pitting
Filiform corrosion
Stress corrosion cracking
Atmospheric corrosion
Oxidation
Figure 12.2 Steam line pinhole leak.JI
• Corrosion testing in the fieldi--30
Corrosion racks
Dutchman • The failure is by brittle fracture.
Electrical resistance probe
Linear polarization resistance • The energy contained in the fluid is large (explosive potential).
Corrosion coupons • The component contains a flashing liquid (discharge ofliquefied gas
or of liquid pressurized and heated above boiling temperature).
12.4 Leak-Before-Break (LBB) • The load, for example, overpressure, is sudden rather than gradual.
Most failures oftanks, vessels, and piping are by leakage through a pin- • The overload is caused by large imposed movement that tears the
hole or tight crack, as shown in Figs. 12.1 and 12.2. Unfortunately, some material.
failures are large ruptures, breaks. A critically important question, • The overload is caused by internai or externai explosion.
when looking at consequence of failure, is to assess when will a failure
be a leak, or when will it be a break? A break is more likely under one It is possible to calculate and quantitatively predict the failure mode
or several of the following conditions: Oeak or break), as illustrated in this chapter.
384 Ch apter Twelve Failure Analysls 385
TABLE 12.1 Stored Energy
Energy
lnitial energy, atambient, Difference,
lnitial condition Btu/lb Btullb Btullb
Water at 50°F and 150 psi 38 (liquid) 38 (liquid) o(liquid-to-liquid)

Water at 340°F and 150 psi 311 (Jiquid) 1124 (steam) 8 13 (liquid-to-steam)
Steam at 400°F and 150 psi 1128 1146 18 (steam-to-steam)
T
P = 150 psi
Figure 12.4 Hydrotest ejection of cap.

Figure 12.3 Energy released at
phase change. when a leak occurs the joint is tightened and the test can continue. In
a few rare instances, when testing with water at high pressure, the joint
v may hold until a high pressure is reached and then give suddenly.
Although water is incompr essible, and the pressure will quickly drop
as soon as the leak forms, there have been instances of severe damage.
12.5 St ored Energy Associated with
Figure 12.4 illustrates the case wh ere a skid was being hydrotested at
Flashing Liquids
high pressure, when- all of a sudden-a threaded male cap carne loose
To understand boiler explosions, or more generally the violent failure and was propelled like a bullet, injuring a nearby worker. The cap is
of equipment containing flashing liquids, we must consider their shown, on the floor, in front of the skid.
stored energy. The stored energy of a fluid inside a vessel or pipe is the Upon investigation it became apparent that the threaded cap was
difference between the energy of the fluid at operating pressure and its corroded, and the threads could not hold the hydrostatic pressure for
energy at ambient pressure. In Table 12.1 we compare three systems: long. The force propelling the cap was
a liquid system, a "superheated" liquid system, and a steam system, ali
three at the same intial pressure of 150 psi. F=PA
The greatest stored energy is present in System 2, where the pressur-
where F = force, lb
ized liquid water operates, pressurized, above 212°F and will change
P = internai pressure, psi
phase by flashing to steam in the event of a leak or rupture. The temper-
A = cap cross-sectional area, in2
ature-volume diagram of Fig. 12.3 helps illustrate the work that the
stored energy produces, W = P dv.aa For example, at 5000 psi, a 4-in2 cap will be propelled with a force
F = 5000 X 4 = 20,000 lb.

12.6 Hydrotest Failure
A rupture in a high-pressure liquid system can be catastrophic even if 12.7 Gas or Liquid Contents
the liquid does not vaporize. It is not uncommon for flange joints or
other mechanical joints to leak during hydrostatic testing. In the vast Once it has been established that a system, equipment, or component is
majority of cases the hydrostatic test pressure is raised slowly, and not fit for service, either because it is significantly degraded or it cannot
386 Chapter Twelve Fallure Analysis 387
sustain an overload, or both, it is often necessary to try to understand

how the component will fail:
• Will the failure be a leak ora rupture?

• How large will the rupture be? Will it be a pinhole leak, a fishmouth,
or a running crack?
These questions are addressed in part in Chap. 9, through the failure

assessment diagram, but we show here the role of the contents on the
failure mode.
Once a failure of pressure boundary takes place, the behavior of the
system, equipment, and component varies significantly depending on
the phase ofthe contained fluid. The difference between gas and liquid
is straightforward: it is the difference between the puncture of an air-
filled birthday balloon and the puncture of the same balloon, but
water-filled. The air is compressible, and will maintain its pressure for
a longer time, propelling the punctured balloon around the room or
violently ripping it to pieces. The water-filled balloon will simply leak
and deflate. Figure 12.5 Tank roof blown off during hydrotest.
This is the reason why it is essential to vent air or gases from a liquid
system before it is pressurized. The pressure is the same in the trapped
air pocket as it is in the surrounding liquid. The onset of failure is the
same in both cases, but, once the failure occurs, the air pocket behaves
as the air-filled birthday balloon whereas the liquid-filled system would,
in most cases, safely leak and drop its pressure instantly.
12.8 The Tank Top Example

In Fig. 12.5, a storage tank was being filled with water for leak test-
ing. The water was piped from a fire hydrant at a higher flow rate than
the tank top vent could handle. Some air remained trapped in the tank
as it was being fliled with pressurized water. The tank roof-to-shell
joint ruptured and the energy in the compressed air trapped in the
tank blew the roof off.
Ifthere were no trapped air, the tank would have been filled with water,
and if the water had reached the same pressure, it would have also
ruptured the roof-to-shell joint, but the energy in the water would have
been insuffi.cient to blow the roof off; instead, the joint would have opened Figure 12.6 Frangible roof design.
locally and spilled some water, immediately dropping the pressure.
contents, or the bottom plate-to-shell weld would have failed, propelling
the tank in the air like a rocket. This is why tank design is based on the
12.9 Tanks with Frangible Roof Design
principie of defense in depth:
The failure mode depicted in Fig. 12.6 is actually preferable to a failure
in which the side walls of the tank would have ruptured, emptying the • First, by system design, prevent gradual or explosive overpressure.
• Second, mitigate overpressure by properly sized vents, rupture discs, 12.11 Leak-before-Break Using the
or explosion panels. Failure Assessment Diagram
• Third, provide a frangible roof design in which the vessel top-to-shell The Failure Assessment Diagram (FAD) is a powerful tool in helping
joint is weaker than the shell (Fig. 12.6). predict whether a failure will occur, and if it does occur whether it is a
leak (pinhole or narrow crack) ora break (gaping crack or separation).
12.1 O Stored Energy The following simple example will help illustrate this point.
The energy stored in an ideal gas is32 8-in schedule 40 pipe (322 mil nominal wall)
ASTM A 53 Grade A carbon steel
E = PV
k-1
[1-(pamb)k-\l
p
P = 325 psi at T = 425°F (saturated steam)
gas Sy = 25 ksi at 425°F (yield stress)
KIC = 100 ksi-.Jin (fracture toughness)
where V = volume of gas, in3 Assume for this example, that the only stress is the hoop stress due
k = gas constant
p amb = ambient pressure, psia to pressure; with primary bending, secondary, and residual stresses
negligible. If the pipe contains a part-through wall crack then
P = pressure of stored gas, psia
The energy stored in a liquid is Case 1. Behind the crack is a ligament of 62 mils remaining wall
1 p2y Kr,62mils = 11 ksi,lin stress intensity --7 Kr.62mils =11 / 100 = 0.11
E--- at crack
L 2 p
S ref,62mits = 23 ksi reference stress --7 Lr,62mils = 23 / 25 = 0.92
where p = bulk modulus ofliquid, psi (330,000 psi for water). This point is inside the FAD, pointAin Fig. 12.7; the pipe does not fail.
For example, the energy contained in a 5-ft3 gas bottle filled with air Case 2. The crack progresses to the point where the remaining wall is
at 2000 psi is 24 mils
E
gas
= 2015 X8640 [1- (~)O.:l-{_4]
1.4 -1 2015
= 33 X106 inOb
Kl.24mils = 27 ksi-.Jin stress intensity
at crack
--7 Kr,24mili = 27 /100 = 0.27
Sref,24mils = 62 ksi reference stress --7 Lr,24mils = 62 f 25 = 2.5

Egas = 2.7 X 106 ft-lb- 1.7 lb ofTNT This point is outside the FAD, in the ductile fracture zone (high Lr
with low Kr), point B in Fig. 12.7; the pipe should fail under these con-
The energy contained in the same bottle filled with water is ditions in a ductile mode. By calculating the location of the projected
point (Lr, Kr) in the FAD diagram, we can predict whether the failure
E = ! 20152 X 8640 - 53 X 103 inClb will be a leak <Iow Kn large Lr) or a break Oow Lr, large Kr), as
L 2 330,000 explained in Chap. 9.
EL = 4430 ft-lb = 0.003 lb TNT References

There is over 500 times more energy in the 2000 psi, 5-ft3 gas-filled 1. Uhlig. H. H., The Corrosion Handbook, Wiley Interscience. New York.
bottle than in the same bottle filled with water. This is the energy that 2. Shrier, L. L., Corrosion, Butterworth, Woburn, MA.
3 ASTM G 1, Standard Practice for Preparing, Cleaning, and Evaluating Corrosion Test
will drive fracture and cause fragmentation. Specimens.
Krr--------------------------, 25. ASTM G 44. Standard Practicc for Exposure of Metais and Alloys by Altemate
Immersion in Neutra[ 3.5 9é Sodium Chloride Solution.
26. ASTM G 36. Standard Proctice for Evaluating Stress· Corrosion·Cracking Re.çistance
of Metais and Alfoys ITI a Boiling .\-fagrlf!sium Chloride Solution.
27. ASTM G 37, Standard Proctice for Use ofMattsson ·s Solution ofpH 7.2 to Emluatc
lhe Stress-Corrosion Crocking Susceptibility of Copper-Zinc Alloys.
28. ASn'l G 50, Standard Proctice for Conductíng Atmo11pheric Corrosicn Tests o" Metais.
29. ASTM B 117, Standard Practice for Operating Salt Spray tFog) Apparatus.
30. ASTM D 1654. Standard Test Method for Evaluaticn ofPainted or Coated Speeimen.~
Subjected to Corrosil•e Enr:ironments.
B 31. Sperko. W. J . Sperko Engineering Services Inc., Greensboro. NC.
32. Antaki, G. A.. Piping and Pipeline EnginHring, Dekker, ~ew York.
A. 33. R. Farish. pen:onal communication, January 2005.
Figure 12.7 FAD for Example
Cases 1 and 2.
4. ASTM G 31, Standard Practice for Laboratory lmmersion Corrosion Testing ofMeta/s.
5. ASTM G 4, Standard Guide for Conducting Corrosion 'Ihts 111 Field Applications.
6. ASTM G 52, Standard Practice for Exposing and Eualuating Metais and Alloys in
Surface Seawater
7 ASTM D 2688, Standard Test Methods for Corrosivity of Water in the Abscnce o{ Heat
T'ransfer !Weight Lo.,, Methods).
8. ASTM G 5, Standard Reference Test Method for Making PotenltMlatic and Potentio·
dwuzmic :'\nodic Polarization Measurements.
9. ASTM G 61, Starzdard 'I'est Method for Conducting Cyclic Potentiodynamic Polari.zation
Measurements fo,. Loca/ized Corrvsion Susceptibility of lron·, Nickel·. or Cobalt·Based
Alloy,ç.
10 ASTM D 2';'76. Stot• rlard Test ,\fethods for Corrosivity ofWater in tire Absencc of Heat
T>·rJn,çfer.
11. ASTM G 59. Sta "lard Test .\fethod for Conducting Potentiodynamic Polarizat10n
Res1stanc.· .\feasurements
12. AST:\1' G 69, Standard Thst Afethod for <lleasurement o{ Corrosion Potentials o{
Aluminum Allm·s.
13. ASTM A 262. St~ndard Proctice.• for Detect1ng Susceptibility to lntergronu/ar Attack
in Ausfellltlc Staudess Steels.
14. AST:\1' G 28. Standard Test .\fethods of Detecting Susceptibílity to lntergronular
Corrosion '" Wrt>ll~ht. Nickel·Rich, Chromium·Bearing Alloys.
15. ASTM G 67, Standard 1bt Method for Detennming the Susceptibility to lntergronular
Corrosion o{5XXX Series Afuminum .4.lfoys by Afass Loss After Exposun· te Nitnc Acid.
16. ASTM G 34, Standard Test Jlethod for Exfoliation Corrosion Suscept1bilit.v in 2XXX
and 7XXX Serie.ç Alumirwm Allovs.
17. ASTM G 66, Standard Test Metl;od for Visual Assessment of Exfoliation Corrosion
Susceptíbility of5XXX Series Aluminum Alloys.
18. ASTM G 48, Standard Test Methods for Pitting and Crevice Corrosion Resistance of
Stainless Steels and Related Alloys by Use of Ferric Chloríde Solut10n
19. ASTM D 2803, Standard Guide for Testing Filiform Corrosion Resistance ofOrganie
Coatings on Metal
20. ASTM G 38, Standard Practice for Making and Using C·Ring Stress-Corrosion Test
Specimen
21. ASTM G 30, Standard Practiee for Making and Using U·Bend Stres.~·Corrosion Tcst
Specimen.~
22. ASTM G 39. Stanrlrml Pmctice for Preparaticn and Use of Bent-Beam Stress-Corrosion
Test Specunens.
23. ASTM G 49. Sta iard Practice for Preparation and Use of Direct Tension Stn:ss-
Corrosion Te.~t SP' ·ímens.
24. AST).l G 58 Sta' f, rrd Proctice for Preparotion of Stress-Corrosion Test Spec:imens
for Weldnumts.
Chapter
13
Repairs
13.1 Repair Work Package

A repair is a miniproject, as such, the first five steps of equipment
integrity apply to repairs (see Chap. 1). Each repair plan and repair
work package should be clear, succinct yet complete, and contain the
following sections.
• Safety
• Codes, standards, procedures, and regulations
• Materiais
• Design
• Fabrication
• Examination
• Testing
In this chapter, we first address considerations common to most
repairs. This is followed by a description of each repair technique.
13.2 Postconstruction Codes and Standards

There are several codes, standards, and regulations that address post-
construction activities for tanks, vessels, and piping, including opera-
tions, maintenance, and repairs. They include:
• ASME B31.1 Power Piping

• ASME B31.4 Liquid Petroleum Transportation Piping
• ASME B31.8 Gas Transmission and Distribution Piping
393
394 ChapterThlrteen Repairs 395
• ASME B31G Manual for Determining the Remaining Strength of An AS.ME postconstruction code is under development. It will address
Corroded Pipe risk-bÇlsed inspections, repairs, and fitness-for-service.
• ASME B31.8S Managing System Integrity of Gas Pipelines
• ASME VI Recommended Rules for the Care and Operatzon of Heating 13.3 Temporary or Permanent Repair?
Boilers Repairs should not be labeled "temporary" or "permanent." Instead, each
• ASME VII Recommended Guidelines for the Care of Power Boilers repair should be assigned a design life, based on a competent analysis of
• ASME XI Rules for ln-service lnspection of Nuclear Power Plants degradation and remaining strength. The repair should be replaced or
upgraded prior to reaching the end of its design life, with margin. This
• 10 CFR Energy, Part 50, Domestic Licensing of Production and margin may be based on one of the following techniques.
Utilization Facilities
• 29 CFR Labor, Part 1910, Occupational Safety and Health Standards • Half-life: For example, if a repair has a design life of four years, it
should be replaced or at least inspected after two years.
• 49 CFR Transportation, Part 192, Transportation of Natural Gas
and Other Gas by Pipeline: Minimum Federal Safety • A practical reliability analysis. 1
• 49 CFR Transportation, Part 193, Liquefied Natural Gas Facilities:
Federal Safety Standards. Part 194 Response Plans for Onshore Oil 13.4 Safety
Pipelines Ali repairs must be planncd and controlled to ensure worker and pub-
• 49 CFR Transportation, Part 195, Transportation ofHazardous Liquids lic safety. These are not idle words as many accidents happen during or
Pipelines shortly after repairs. The repair package should be reviewed by all
• API 510 Pressure Vessel Inspection Code: Maintenance, lnspection, responsible parties to ensure, as a minimum, the adequacy of protective
Rating, Repair, and Alteration gear, breathing ai r where required, emergency response planning, appro-
priate lockout and tag-out, welding exclusion zones, safety of welding on
• API 570 Piping Inspection Code: lnspection, Repair, Alterations, and
a component that has been in serv1ce and may be contaminated with
Rerating of In-Service Piping Systems
products and degraded, confined space entries, general area condition
• API 572 lnspection of Pressure Vessels and restrictions, and sign-offs by operations and the safety engineer.
• API 573 lnspection of Fired Boilers and Heaters
• API 574 Inspection of Piping, 'Iitbing, Valves, and Fittings 13.5 Regulatory Requirements
• API 575 lnspection of Atmospheric and Low Pressure Storage Tanks Some operating companies are required 'or have voluntarily elected)
• API 576 lnspection of Pressure Relieving Devices tD obtain jurisdictional or third-party approval of certain types of
repairs. The responsible repair engineer should factor the applicable
• API 579 Fitness-for-Service
regulations into the repair work package. In the United States, these
• API 580 Risk-Based Inspection would include:
• API 581 Base Resource Document- Risk-Based lnspection
• For ASME pressure vessels and boilers, the Authorized Inspector
• API 598 Valve Inspection and Test oversight and repair stamps (R stamp for a vessel, VR for a safety
• API 653 Tank lnspection, Repair, Alteration, and Reconstruction Code relief valve)
• API 11V7 Repair, Testing, and Setting Gas Lift Valves • For ali pressure retaining items, the National Board's ANSI-NB-23
• API 2201 Safe Hot Tapping Practices in the Petroleum & Petrochemical • For transmission and distribution hydrocarbon pipelines, the Code
Industries of Federal Regulations 49 CFR. and ASME B31.4 and B31.8
• NBIC, National Board Inspection Code, ANSI I NB-23, the National • For nuclear power plants, the Code of Federal Regulations 10 CFR
Board of Boiler and Pressure Vessel Inspectors 50, and ASME XI
396 ChapterThlrteen Repalrs 397
• For process systems containing threshold quantities of hazardous For example, a mechanical clamp is forged from mild carbon steel,
material~. the Code of Federal Regulations 29 CFR 1910 and is meant to repair a pipe with a design pressure of300 psi at 300°F.
• For refineries and petrochemical plants, API 510 and 570 A prototype clamp is tested up to 1600 psi, without damage. Given
S .. = 60,000 psi from the material specification

13.6 Common Considerations for Materiais
s!J..aVg = 70,000 pSÍ ITOm tensile teStS On test SpecimenS CUt OUt frOffi
Materiais for repairs should comply with all the requirements of the the tested part
design and construction code. The materiais selected must be listed in E = 1 because there are no welds
the design and construction code, or, ifproprietary, the materiais should
conform to a written material specification, which includes quality man- f= 1 because the part is forged, not cast
ufacturing, controls on chemistry, physical and mechanical properties, Sdesign = 20,000 psi from the ASME B31 design code, at 300°F
with reasonable quality assurance. For example, a metallic repair sleeve Sroom = 20,000 psi from the ASME B31 design code, at room tem-
should follow an ASTM or ASME II material specification. An epoxy- perature
impregnated carbon fiber should have a manufacturing specification,
and the manufacturing process should be auditable by the purchaser. The assigned pressure rating of identical clamps is therefore
13.7 Common Considerations for Design p = 1600 X 60,000 X l X

1 X 20,000 = 342 Í
rat -d 4 70 000 20 000 ps
The design of a repair should comply with all the requirements of the , '
design and construction code. The pressure rating of repairs should be The clamp can therefore be used to repair a pipe with a design pressure
at least equal to the rating of the original design. The pressure rating up to 342 psi at 300°F.
should be established either by calculation for simple shapes such as The second option for pressure rating a pipe repair component is ASME
cylinders, spheres, and nozzles, or by prooftesting. Prooftesting should B16.9.2 In this procedure, the repair component or fi.tting is required to
comply '>Vith the rules o f the design code. For example, in a proof test in sustain a test pressure equal to
accordance with ASME VIII DiY.1 UG-101 the partis subject to steadily
increasing pressure, preferably hydrostatic rather than pneumatic for 2St
safety reasons. Once a maximum test pressure B is reached without B=1.05
D
damage, the test can be stopped and a pressure rating is assigned to
the part, or an identical part. by the following formula where B - maximum pressure sustained satisfactorily during test, psi
S = actual tensile strength of the tested fitting, psi
B S;;. Sdesígn
P teci =-x--xExfx-- t = nominal wall thickness of the matching pipe, in
.-a 4 s;;.avg smom
D = outside diameter of the matching pipe, in
where Prnted = assigned pressure rating, psi Another, simpler, way to apply ASME B16.9 is to write that the repair
B maximum pressure sustained satisfactorily during component or fitting should be capable of sustaining at least 3 times (for
test, psi B31.3,3 ) or 3.5 times (for B31.1,4 ) the design pressure; and that the repair
sf' - specified minimum tensile strength at room tempera- component's material must have a similar allowable stress as the pipe at
ture, psi design temperature.
s~ J = average actual tensile strength of tested specimen at
room temperature, psi 13.8 Common Considerations for
Fabrication- Welding
f= casting quality factor for cast parts
sd rn = code allowable stress of material at design temperature, The fabrication requirements in the repair package will depend on
psi whether the repair is a welding or a nonwelding repair. For a welding
S~ , = code allowable stress at room temperature, psi repair, the repair weld procedure and welder qualification should comply
398 Chapter Thirteen Repalrs 399
with the construction code, with particular attention to welding on line,

and welding on a contaminated component, topics that have been
addressed in Chap. 4.
A unique difficulty arises with repairs that require postweld heat
treatment. A weld repair should be postweld heat-treated ifthe original
construction weld was postweld heat-treated. Also, a weld repair should
be stress relieved ifthe weld deposit is large, causing significant cooling
shrinkage.
In practice it may be quite a challenge to heat-treat in the field a
vessel that has been in service, much more difficult than heat-treating
the vessel new, often in the fabrication shop. For practica] reasons it
may not be possible to uniformly heat-treat the vessel, instead, heat
treatment is confined to the repaired area. This is a bull's-eye heat
treatment. This nonuniform heating of a vessel has caused vessels to
visibly deform, and accumulate more residual stresses than before the
repair. There are options to heat-treating a weld repair. These options
are desc1ibed, for example, in ANSI-NB-23, and rely on preheat and
controlled deposition welding.
Figure 13.2 Cracks initiated at repaired nozzle.õ
The catastrophic failure of the vessel shown in Figs. 13.1 and 13.2 was
due to a poor repair of a nozzle forging. The manufacturer repaired the
nozzle by weld overlay, \vith two shortcomings:
• The repair was made without preheat, causing subsurface cracks.
• The repair modified the nozzle opening from a rounded corner to a
square corner, causing stress concentrations in service.
13.9 Controlled Deposition Welding

Coarse grains form between weld passes, at locations indicated m Fig. Fig. 13.3 Coarse grain between passes.6
13.3. Controlled deposition welding are welding techniques that rely on
one weld pass heat treating the previous pass, in place of a separate
postweld heat treatment. An example of controlled deposition is temper
bead welding in which each weld pass heat-treats and refines the previ-
ous weld pass, to achieve similar benefits as postweld heat treatment.
As a cost-effective alternative to replacement, a 72-in long, through-
wall crack along the outer are of a 10-year-old, 15-in diameter X
3.375-in thick, 2Y.Cr-1Mo main steam line in a fossil power plant,
operating at 1005°F and 3550 psi (Fig. 13.4) was successfully repaired
by welding. 5
An example of control deposition weld repair of a longitudinal weld
seam is shown in Fig. 13.5. Another example is illustrated in Fig. 13.6,
where a thick section was preheated and welded with a prequalified,
Figure 13.1 Cat.astrophic failure of vesseJ.ô control bead deposit technique to achieve the same final microstructure
400 Chapter Thlrteen Repairs 401
Figure 13.5 Weld repair of longitudmal seam weld.s
Figure 13.4 Cracked pipc prior to repair.6
and hardness as a postweld heat-treated section. The key paramete1·s

for this particular temper bead repair are:
Preheat 250°F Figure 13.6 Temper bead control deposition weld repair.s
Interpass 400°F
Electrode E8018 B2L CAWS A5.5) 13.1 O Postconstruction Standards for
Bead overlap First 3 layers 50 percent bead overlap Controlled Deposition
Cavity welding Start center of cavity going out to edge Postconstruction standards address controlled weld deposition tech-
Electrode angle 90 degrees to bottom of repair cavity niques as altematives to postweld heat treatment, and provide detailed
Welding 1st layer ~ in electrode, 80 A, 10 inlmin, 9 KJ/in requirements for their implementation. These postconstruction stan-
Welding 2nd and 3rd ~ in electrode, 130 A, 10 in/min, 17 KJ 1Ín
dards include:
Remaining layers ~ in and -~ in • API 510 (pressure uessels). "Preheat and controlled deposition weld-
Final temper layer Overfill one layer, finallayer not at base metal. grind ing ... may be used in lieu ofpostweld heat treatment (PWHT) where
out final layer PWHT is inad,ásable or mechanically unnecessary."
402 Chapter Thirteen Repairs 403
• API570 (piping) . "Preheating to not less than 300°F (150°C) may be The practical difficulty when examining welded repairs is the extent
considered as an alternative to postweld heat treatment (PWHT ) for of examinations. Many construction codes apply a percentage approach
alterations or repairs of piping systems initially postweld heat treated to inspections, for example, inspection of20 percent ofwelds. This per-
as a code requirement." centage is meant for a large quantity of welds, for example, a whole
• ASME XIIWA 4610 (nuclear power plants). "Repair/replacement project or a whole new system, but is difficult to apply to a repair that
activities on P-Nos. 1, 3, 12A, 12B, and 12C7 base materiais and may entail only a handful ofwelds. To address this point, keep in mind
associated welds may be performed without the specified postweld the primary objective of exarninations: with welder certification and
heat treatments, provided the requirements of IWA-4621(b) and (c) weld procedure qualification, both the person and the process have
and IWA-4622 through IWA-4624 are met." proven to be competent. So what is in question, what examination will
help flush out is (a) the effect of field conditions on weld quality, and
• NBIC NB-23. When postweld heat treatment is "inadvisable or imprac- (b) the welder's performance at that particular time. Often, a sampling
tical," the National Board Inspection Code provides alternative technique is sufficient to achieve these objectives. For criticai applica-
welding methods, each applicable to specific metais and welding tions, where a leak would be intolerable, 100 percent examination would
processes. In the first method, the metal is preheated to a minimum be in order inasmuch as repairs do not involve many welds.
of 300°F, and the interpass temperature is limited to a maximum of
450°F. The second welding method is a controlled-deposition, temper
bead or half-bead technique, with postweld hydrogen bakeout at 13.13 Common Considerations for Testing
450°F, and notch-toughness test of the qualification weld. The third This is probably where the construction code rules are most difficult to
welding technique calls for a buttering layer, with 50 percent over- apply to repairs. It is sometimes very difficult to pressure test (hydro-
lap stringer beads, and subsequent layers that do not contact the statically or pneumatically) a repair. This is because the repaired section
base metal, also with a final hydrogen bakeout. Bulls-eye PWHT is would be difficult to isolate. The options to solve this problem are:
permitted with limits such as a temperature gradient not to exceed
250°F per foot. • Ad hoc isolation through temporary plugs, such as freeze plugs (Fig.
13. 7) or inflatable plugs or discs
• ASME VII Recommendedguidelines for the care ofpower boilers. This
ASME guide espouses the National Board Inspection Code. • Sensitive leak testing. for example, helium with sniffer or air with bub-
ble solution, as described in Chap. 4
13.11 Common Considerations • Vacuum box test, as described in Chap. 4
for Fabrication-Nonwelding • In-service leak test with augmented examination
For nonwelding repair, the emphasis should be on following vendor
requirements. For example, the vendor catalog for a swaged fitting
may specify that the fitting has to be twisted till it is hand-tight, and
then torqued beyond hand-tight, with a wrench, for one anda quarter
turn. Exacting installation requirements also apply to bolted, wrapped,
or sprayed forros of repairs. The vendor requirement should be followed
by a trained installer.
13.12 Common Considerations

for Examination
This is the quality control step of the installation process. For welding
it would entail nondestructive testing; for nonwelding repair it would
rely on procedural controls of the fabrication and installation process
and visual examination of the completed assembly. Figure 13.7 Freeze plug.7
404 Chapter Thlrteen Repalrs 405
The postconstruction codes address the question of pressure or leak rernoval, is at least 90 percent of the minimum design wall thickness;
testing repairs, as follows. components or connections NPS 1 DN 25 and smaller; tube-to-tube
sheet welds when such welds are rnade on the cladding; seal welds;
• API 510 5 (pressure vesselsJ. ~\Vhen the authorized pressure vessel welded or brazed joints between nonpressure retaining items and the
inspector believes that a pressure test is necessary or when, after pressure-retaining portion of the cornponents; valve discs or seats."
certain repairs or alterations, the inspector believes that one is nec-
essary. the test shall be conducted at a pressure in accordance with
the construction code used for determining the maximum allowable 13.14 Common Considerations
working pressure.'' for Quality Control
• API 5108 (pressure vesse/s). ~subject to the approval ofthejurisdiction When planning tora repair. the facility relies on good quality control in
(where the jurisdiction's approval is required). appropriate nonde- the supply chain of"MRO rna~rials" (maintenance, r epairs, and opera-
structlve examínations shall be required where apressure test is not tions materiais), from procurement to installation. This quality chain,
performed. Substituting nondestructive examination procedures for the assurance that what is installed is what was intended, becomes
a pressure test after an alteration may be done only after a pressure irnportant at various leveis:
vessel engineer experienced in pressure vessel design and the autho-
rized pressure vessel inspector have been consulted.'' • For process-safety cntical applications under safety regulations such
as OSHA, 29CFR1910
• API 510" (pressure vesse/s). "After welding is completed, apressure
test ... shall be performed if practical and deemed necessary by the • For business criticai processes
inspector. Pressure tests are normally required after alterations and • For excellence in conduct-of-rnaintenance
major repairs. When a pressure test is not necessary or practical,
1\TJJE shall be utilized in lieu of a pressure test. Substituting appro- For criticai applications, the quality control should occur at four leveis:
priate NDE procedures for a pressure test after an alteration or repair
may be done only after consultation with the inspector and the piping • Quality control ofthe plant rnaintenance and replacement part data-
engineer." base
• API 57011 (piping). "The owner/user shall specify industry-qualified • Quality control of manufacturers, suppliers, fabricators, and dis-
UT shear wave examiners for closure welds that have not been pres- tributors
sw·e tested." • Quality control at plant receipt inspection
9
• API 2201 (lzot tapping). ''If the current temperature of the line or • Quality control when the technician pulls out the material from stores
vessel will permit, conduct a hydrostatic test ofthe welded attachrnent for use
and hot tapping rnachine in accordance with the applicable code."
• API 2201 9 (hot tapping). "If the temperatme is such that a hydro- 13.15 Replacement
static test cannot be conducted, air or nitrogen with soap solution on
the weld can be used." The rnost cornrnon forrn of repair is to cut out, remove, and replace th e
degraded section. In the case of a pipe, this is generally straightfor-
• ASME X/ 10 IWA 4540 (nuclear power plant componcnts). "A systern ward. In the case of large pressure vessels and storage tanks, sections
leakage test shall be performed in accordance with IWA-5000 prior to are r eplaced as "window" repairs, which are addressed in the next
oras part ofreturning to service. The following are exempt frorn any section.
pressure test: cladding; heat exchanger tube plugging and sleeving; The design rnay be rnodified to either eliminate the cause ofthe repair
piping, pump, and valve welding or brazing that does not penetrate or facilitate future repairs ofthe sarne section. This is, for example, the
through the pressure boundary; flange seating surface when less case illustrated in Fig. 13.8, where a welded tee that was r epeatedly
than halfthe flange axial thickness is removed and replaced; pres- repaired was finally replaced by a flanged spool easier to replace,
sure vessel welding when the remaining wall thickness. after metal because the cause of degradation could not be eliminated.
406 Chapter Th lrteen Repairs 407
In API 510, the following conditions are specified for flush patch repairs
of vessels:8
• Material, design, fabrication, inspection, and testing per code.

• Patches shall also have rounded corners.
• Patches shall be installed with full-penetration butt joints.
13.17 Example of Flush Patch Repair

A 70-ft tall tower, 96-in (8 ft) in diameter X 3.875-in wall, ASTM A 516
Grade 70 N, with a design pressure of 1300 psi at 200°F, had developed
hydrogen sulfide induced cracks (HIC). The flush patch repair option
included the following steps.
• UT to delineate HIC zone

Figure 13.8 Flanged tee to ease
replacement • HIC resistant plate with limits on carbon eq uivalent
• UT the replacement plate
• Hydrogen bake-out the replacement plate and repair area 600°F for
1.5 h/in thickness
13.16 Flush Patch Repair
• WFMPI' (wet fluorescent magnetic particles) on weld bevels
A pressure vessel or storage tank can be repaired by cutting out a win- • Qualify weld procedure for hardness of weld-HAZ
dow that contains the defect to be eliminated. and welding in place a
curved plate. butt-welded flush to the existing shell or head. The insert • Preheat 300cF, maximum interpass 450°F, E7018-1 low hydrogen
plate may be round, oval, or square with rounded corners. This repair is electrode
also used to add a nozzle to a vessel, tank, or pipe, or to access the back • Weld one side then back-gouge root and inspect with PT
wall of a boiler tube. The materiais, design, and construction rule:; for • Complete welding, tJT or RT finished weld
the flush patch should comply with the construction code. Special pre-
cautions are in order when the new patch intersects an existing weld: • PWHT 1175°F ± 25()F for 2 h
• Hydrotest 1.5 X design pressure
• The existing weld should intersect the new flush weld ata wide angle
of 30 degrees or more. • After hydro WFMPT
• The existing weld should be exarnined before and after welding the
flush patch to rnake sure that the repair weld did not cause the exist- 13.18 Flaw Excavation
ing weld to crack. A base metal or, most often, a weld defect can be excavated by thermal
gouging (carbon are or plasma are) or by mechanical grinding (grinding,
Insert flush patches for piping repairs are addressed inAPI 570, and
honing; Fig. 13.9). The gouged or ground surface can then be inspected
approved under the following conditions. 11
with liquid penetrant to confirm that the flaw has been completely elim-
• Full-penetration groove welds are provided. inated. When thermal gouging, the therrnally affected metal should be
finished by rnechanical grinding to a depth of at least 1/16 in to elirni-
• One hundred percent RT or UT of criticai welds. nate the heat-affected metal. Note that grinding a severely thinned wall
• Patches rnay be any shape but shall have rounded corners, \vith at has a good chance of causing a growing crack. This will eventually
least a l-in comer radius. require replacement of the part.
408 ChapterThirteen Repairs 409
The component wall thickness should then be restored to the design

code required minimum thickness plus future corrosion allowance by
deposition welding (Fig. 13.10). The repair cavity should be kept as
narrow as possible to minimize weld residual stresses due to shrink-
age, which could even lead to crack.ing at the periphery of the repair
weld.
If there is sufficient remaining thickness after metal remova!, the
cavity may be left as-is, provided it has a smooth contour and is away
from geometric stress risers. A cavity contour with a 3:1 slope is usu-
ally sufficiently smooth to prevent stress concentrations.
Once the repair weld is complete, including postweld heat treatment
where required, the weld repair should be visually inspected (VT) and
the finished surface should be PT or MT for evidence of shrinkage or
Figure 13.9 Grinding out a weld defect. hydrogen cracking. If the excavation has fully penetrated the wall,
then it may be necessary to RT or UT the completed weld repair.
13.19 Example of Flaw Excavation Repair

A 70-ft tall tower, 96-in (8 ft) in diameter X 3.875-in wall, ASTM A 516
Grade 70 N, with a design pressure of 1300 psi at 200°F, had developed
hydrogen sulfide cracks. The flaw excavation option included the fol-
lowing steps.
• UT to delineate HIC zone.

• Grind out crack.
• Hydrogen bake-out repair area 600°F for 1 h/in thickness.
• Preheat 300°F, maximum interpass 450°F, E7018-1 low hydrogen
electrode.
• Temper bead procedure API 510 or NBIC.
• Ensure toe weld-base metal is not last pass. Grind out final temper
bead pass.
• Maintain 500°F postweld for at least 2 h.
• WFMPT inspection of repair.
• Check weld and HAZ hardness < 200 Brinell.
13.20 Weld Overlay
Figure 13.10 TIG deposition in Figure 13.11 illustrates the weld overlay repair of a longitudinal crack,
repair cavity. 50 percent through-wall. The original wall thickness tw is covered with
two dilution layers of total thickness toL, and then the third weld over-
lay layer of thickness twoL with
410 ChapterThirteen Repairs 411
' +~tD.,L"------
_ _t_,_,"_ = 0.75 13.21 Full Encirclement Welded Sleeve
t\1' + fDL + fl\"OL Welded sleeves have been used for decades to repair oil and gas
pipelines. 13- 15 They consist of two bent plates tightly fit around the
The original crack is conservatively assumed to remain embedded, pipe (Fig. 13.131, and welded longitudinally (type A sleeve) or also
and to have progressed into the two dilution layers. Under these condi- welded circumferentially to the degraded pipe at the two sleeve ends
tions, a detailed crack stability and fatigue analysis is performed to (type B sleevel Multiple sleeves may be used to repair long defects
verify the integrity ofthe repair and establish a design life ofthe repair. (Fig. 13.14).
Figure 13.12 shows a welder qualification for weld overlay repair. When fully welded longitudinally and around the circumference the
sleeve provides two benefits:
• It compresses the degraded pipe, preventing it from bulging outward
at a thinned wall section.
• It contains an eventual leak.
Figure 13.11 Weld overlay repair, thírd layer.l2
Fig.13.13 Sleeve clamped in posí-

tion hef- re welding.
Figure 13.12 Weld overlay qunlífication.7 Figure 13.14 Multiplc slecves repair.;
412 ChapterThlrteen Repairs 413
The sleeve is also commonly used to repair dents in pipelines. The • The pressure should be reduced before installing a sleeve. Standard
dented area is first filled with a hardenable resin that is shaped to practice is to reduce the pressure by 20 percent.
the cylindrical contour by grinding, and then sleeved. The sleeve stiff-
ness will bear down on the pipe and prevent the dent from breathing in Full encirclement sleeves are also a vessel repair technique, in which
and out as the pressure fluctuates. This will, ifwell executed, eliminate case they are referred to as full encirclement lap bands. The conditions
the risk of fatigue rupture at the dent. of applicability in API 5 10 are: 8
Sleeves are recognized as repair methods in several standards,
• Nota crack repair method.
including API 57011 and ASME B31.4. 13 In API 570 the technique is
labeled a "temporary repair," but refer to Sec. 13.3 regarding this point. • The band alone can cont ain the full design pressure.
Some cautions specific to sleeve repairs are: • Longitudinal seams in the repair band are full penetration butt
welds.
• Longitudinal cracks may run and propagate from under the sleeve.
• The weld joint efficiency and inspection are consistent with the code.
• The sleeve material for the repair shall match the base metal.
• The circumferential fillet welds attaching the band to the vessel shell
• The gap between the sleeve and pipe should be very small, on the order are designed to transfer the full longitudinalload in the vessel shell,
of 3/32 in, prior to welding. using a joint efficiency of 0.45, without credit for the vessel shell.
• The longitudinal welds should be completed before the circumferential • Fatigue is considered in the design of the attachment welds.
welds.
• The band material and weld metal are suitable and compatible.
• The sleeve should have a thickness equal to 1.5 times the nominal
thickness of the repaired pipe, with a fillet weld leg approximately • Plan on future inspections, if necessary.
equal to the wall thickness plus any gap sleeve-pipe.
• For short flaws, L< (20 D t)0 ·5 , the sleeve may be sized with a thick- 13.22 Welded Leak Box
ness of0.75 to 1.0 times the pipe wall thickness, ifjustified by fitness- A welded leak box consists of an enclosure, in a variety of shapes, used to
for-service analysis. seal off a degraded component. Normally, leak boxes are used to contain
• A compression repair sleeve involves preheating the sleeve just prior
to installation.
• The sleeve can be shaped to bridge a protruding gir th weld (Fig.
13.15).
• Keep in mind that in case of leak the fluid will be trapped under the
sleeve which could lead to crevice corrosion, and will apply an externai
p ressure to the pipe.
Figure 13.15 Welded split sleeve

for girth welds.
Figure 13.16 Leak box repair oftee.l6

414 ChapterThlrteen Repalrs 415
'I
, - "'
I '
I
I
- _,: I
_. , ~,
Figure 13.17 Leak box repa1r of straight pipe.
leaks at packings, and at flange and gasketed joints, or to contain leaks

(or potentialleaks) dueto local thinning. The leak box may not prevent Figure 13.18 Small fillet-welded patch repair.
the propagation of a crack in the pipe or component. Keep in mind, that
ihe box could cause the following:
• Crevice corrosion in the annulus box-component
• In gas service, corrosion due to condensation of leaking gases
• lfboxing a leaking flange, externai corrosion ofbolts by process fluid
temperature
13.23 Fillet-Welded Patch

A patch plate is fillet welded to cover the arca to be repaired (Fig.
13.18). lts purpose isto contain leaks, and reinforce the degraded wall.
This is one of the simplest repair techniques. The patch plate thick-
ness typically varies from 1.0 to 1.5 times the nominal wall of the
repaired component. If the wall being repaired is too thin, it will tend
to crack when welding the patch plate (Fig. 13.19).
For piping. the repair method is addressed in API 570, and pennitted
Figure 13.19 Patch is too thick.
under the following conditions:11 pipe too corroded.
• The repair area is localized (e.g., pitting or pinholes).

• The yield strength of the pipe is not more than 40 ksi.
• The repair is temporary (but see Sec. 13.3). • They shall be approved by thejurisdiction and the authorized pressure
• The patch is properly designed. vessel inspector.
• The material for the repair matches the base metal unless approved • The fillet-welded patches provide design safety equivalent to rein-
by engineering design. forced openings.
• The flilet-welded patches are designed to absorb the membrane
The following conditions apply to externai or interna! flilet-welded strain of the parts.
8
patches on vessels, according to API 510:
• The stress in the fillet weld is within the code allowable limit.
• They may be used to make temporary repairs (but see Sec. 13.3). • Overlay patches shall have rounded comers.
416 ChapterThirteen Repalrs 417
Fillet welds in large patches on large pressurized surfaces will be sub- • The patch shall not exceed 48 in X 72 in, but should be at least 4 in X
ject to significant shear stress if pressurized as a result of leakage. The 4 in.
rninimum weld size Wmm required to resist the shear stress due to pres-
sure is • The areas to be welded shall be ultrasonically inspected prior to weld-
ing the patch.
PD • Repair plates shall not be lapped onto lap-welded shell seams, riveted
F.c
. =-
2 <wrrun
- xE·Sn
shell seams, other lapped patch repair plates, distorted areas, or unre-
paired cracks or defects.
• The welding shall be continuous aiong the perimeter.
where Fc= circumferential force per unit length, lb-in

13.24 Mechanical Clamp
FL = longitudinal force per unit length. lb-in
P = internai design pre~sure, psi Amechanical clamp is made oftwo haives bolted together. The perime-
D '"" midwall diameter of vesseL in ter of the clamp is bordered by a gasket. The clamp is installed around
Wmin = minimum size of weld leg (si de, not throat), in a degraded or leaking pipe and bolted in place (Figs. 13.20 and 13.21).
E = 0.50 weld joint e fficiency factor In some cases the clamp alone is capable of confining the leak. In oth-
S a allowable stress of weld metal, psi er cases a sealant is injected through a nozzle inside the clamp; the
sealant solidifies and stops the leak.
An option to reduce fillet-wcld stresses is to add plug welds to the :Mechanical clamps are often available as catalogue items or they
pakh, around its circumfercnce The plug hole diameter D would be can be custom-made of two half-shells, a sealing gasket, and bolts or
obtained follo\\--ing the rules of AISC .\!anual of Steel Construction studs and nuts. The assembly is usually designed and fabricated to the
rules of the pressure vessel code or the B31 piping code.
The clamp may not prevent the propagation of a crack in the pipe or
D = 1.3 {E +0.25 in component. Therefore, leak clamps shall not be used when cracks are
vs:: present. The same cautions as for welded leak boxes apply to a
where P = shear on plug weld, lb and D = bottom diameter of plug mechanical clamp.
weld hole. in, with the additional condition 2t + 0.25 in 2::: D <::: t + 0.25
in where t = vessel wall, in.
Storage tankB can be repaired with lap-welded patch plates, under a
long list ofconditions which include17 (refer to API 653 for full conditions
ofrepair):
• Materiais shall meet code.

• The repair should not be used on tank walls exceeding ~ in.
• The patch thickness shall be calculated using dcsign code rules with
ajoint efficiency of0.70 maximum, and should not exceed max r:Xnin;
min (Lmeu;~ in 1].
• The patch may be circular, oblong, square. or rectangular.
• Ali corner:;, except at the shell-to-bottom joint. shall be rounded to a
minimum radius of 2 in.
• The patch may cross shell seams that have been ground flush, but
must O\erlap at least 6 in beyond the shell seam. Flgure 13.20 Mechanical clamp repair.
joined, restoring the system. Vent risers or drains can be installed to

detect leaks to the liner-pipe annular space. 19
In an alternative technique a liner is folded into a C-shape and then
inserted into the degraded host pipe. In yet another method, a resin-
impregnated felt tube is inverted into a degraded host pipe, folding
inside-out as it progresses through the host pipe, and creating a new
corrosion-resistant, load-carrying tube within the pipe.20 These repairs
are particularly efficient for underground utility or process piping
because they are practically trenchless repairs. 21- 23
13.26 Pipe Splitting

A degraded underground concrete or cast iron pipe can be split in place and
replaced by a new polyethylene pipe. This is achieved by first accessing the
pipe at two points: either two existing manholes or by excavation and cut-
ting the pipe. A bullet head with steel blades (Fig. 13.23) is introduced at
Fig. 13.21 Mecbanical clamp witb injection.18 one end and pulled by cable towards the other end. 'Ib the back ofthe bul-
let is attached a polyethylene pipe. As the bullet is pulled through the
existing degraded pipe, it splits the pipe and pushes it aside, allowing
the polyethylene pipe to take its place. This operation is accomplished in
one shot. If there are branches to the replaced header they would have to
be re-established. Road and rail crossings will have to be evaluated because
the compaction has been affected by the rupture of the existing pipe.
13.27 Sacrificial Component

The bend in Fig. 13.24 is subject to significant erosion at the extrados
from fly ash. It is fabricated of two bolted halves, and the top half, the
extrados that erodes, is periodically replaced, before it wears through
the wall.
13.28 Nonmetallic Wrap
Figure 13.22 Pulled liner. Repair laminates are wrapped around the defect and then cured to
form a tight pressure boundary. There are several types of nonmetal-
lic wraps, which include:
13.25 lnserted Liner • Fiberglass tape with water-activated polyurethane resin25·26

• Carbon fiber weave, impregnated with epoxy resins, chemically
A pipe that had suffered severe internai corrosion was repaired by an activated18
insertion liner (Fig. 13.22). The pipe was cut in 300-ft sections, and fitted
with flanges welded at the ends of each section. The pipe was internally
inspected and cleaned. A nylon liner with the right resistance to perme-
ation, chernical resistance, and strength was stretched to slightly reduce
its diameter and then pulled through each section. The liner was then
secured to the flange face at the end of each section, and the flanges Figure 13.23 Pipe-splitting bullet bead.24
420 ChapterThirteen Repalrs 421
Fíg. 13.24 Sacrificial bend component.
Figure 13.26 Carbon fiber wrap repair. 18
Figure 13.25 Carbon fiber wrap repair of elbow. 1~
Rather than a simple wrap, these repairs should be viewed as a system Figure 13.27 Reinforccmcnt of defective welds.18
comprised ofthe component being repaired, its surface preparation, the
nonmetallic composite. which in turn is compri!>ed of a weave or lami- In criticai applications, where the consequence ofleakage or rupture
nate, filler material, and an adhesive. is unacceptable, the nonmetallic wrap should be prequalified. The
Each wrap system is characterized by: parameters to be qualified include:
• A design pressure and temperature limit. These are established by • Short-term mechanical properties. Modulus (ASTM D3039), shear
proof testing and are a functíon of the number of wrap layers. For modulus <ASTM D5379), bending modulus <ASTM D790), expansion
example, an 8-layer \\Tap may have a rupture-leak pressure of 600 coefficient (ASTM E83ll, glass transition ternperature (ASTM E831
psi with a l-in diameter hole. Therefore, this repair may be used as and ASTM El640). and hardness (ASTM D2583).
an unlisted component to 600/3 = 200 psi design pressure.
• Qualzfication test. A defect machined into the pipe wall and_ repaired
• Specified physical properties, such as stiffness in tension and bending. with the wrap system shall be capable of sustaining a hoop stress
• A design life. equal to yield in the nominal wall, away from the machined defect.
2. ASME B16.9, Factory-Made Wrought Steel Buttwelding Fittings, American Society of

3. ASME B31.3, Process Piping, American Society ofMechanical Engineers, New York.
4. ASME B31.1, Power Piping, American Society of Mechanical Engineers, New York.
\.
5. Thielsch Engineering, Cranston, RI.
6. WRC 407, Reports on Heat Treatment of Steels Used in Boiler and Pressure Vessel
Applications: (1) Carbon Migration in Cr-Mo Weldments Effect on Metallurgical
Structure and Mechanical Properties, by C. D. Lundin, K. K. Khan, and D. Yang and
(2) ASME Post-Weld Heat Treating Practices: An lnterpretive Report, by C. E. Spaeder,
Jr. and W. D. Doty, December 1995.
7. Rosenfeld, M., Recommended practices for pipeline repairs, in lnternational Pipeline
Conference, Calgary, October 2004.
8. APJ 510, Pressure Vessel lnspection Code: Maintenance lnspection, Rating, Repair,
and Alteration, American Petroleum Institute, Washingt4Jn, DC.
9. API RP 2201, Safe Hot Tapping Practices in the Petroleum & Petrochemicallndustries,
10. ASME Xl,ln-Service lnspection, American Society ofMechanical Engineers, New York.
11 API 570, Piping lnspection Code lnspection, Repair, Alteration, and Rerating of ln-
Figure 13.28 Fiberglass wrap of elbow-support fillet-weld.
Service Piping Systems, American Petroleum Institute, Washington, DC.
12. Structural Integrity As84lciates, Inc, CA.
13. ASME B31.4, Pipeline Transportation Systems for Liquid Hydrocarbons and Other
Liquids, American Society of Mechanical Engineers, New York.
14. ASME B31.8, Gas Transmission and Distribution Piping Systems, American Society
of Mechanical Engineers, New York.
15. AGA, American Gas Association, Pipeline Repair Manual, December 1994.
16. Mair, D., Design and remnant life analysis of a carbon steel containmentjacket oper-
ating in the creep regime, in Proceedings of PVP2004, ASME Pressure Vessels and
Piping Conference, July 2004.
17 API 653, Tank lnspection, Repair, Alteration, and Reconstruction, American Petroleum
18. Citadel Technologies, Thlsa, OK.
19. United Pipeline Systems, Durango, CO.
20. Insituform Technologies, Chesterfield, MO.
21. ASTM F 1867, Standard Practice for lnstallation of Folded/Formed Poly (Vinyl
Chloride) (PVC) Pipe Type A for Existing Sewer and Conduit Rehabilitation, ASTM
International, West Conshohocken, PA.
22. ASTM F 1871, Standard Specifi.cation for Folded I Formed Poly (Vinyl Chloride) Pipe
Type A for Existing Sewer and Conduit Rehabilitation, ASTM International, West
Figure 13.29 Proof test of fiberglass wrap repair.27 Conshohocken,PA.
23. ASTM F 2207, Standard Specifi.cation for Cured-in-Place Pipe Lining System for
Rehabilitation of Metallic Gas Pipe, ASTM International, West Conshohocken, PA.
• Adhesion strength. Lap shear (ASTM D3165); an average lap shear 24. McElroy Inc., 'fulsa, OK.
strength of about 600 psi should be expected. 25. InduMar Products, Inc., Houston.
26. Neptune Research, Inc., West Paim Beach, FL.
• Long-term performance. Integrity under pressure, either for 1000 h 27. Clock Spring Company, LP, Houston.
(42 days), or by regression analysis (ASTM D2992); or long-term lap
shear, for example, after 1000 h water immersion.
On hot systems, the repair should not be used at a temperature in

excess of the glass transition temperature minus 50°F (ASTM E831,
ASTM E 1640).
References
1. Geitner, F. K., Setting inspection frequencies, Pipeline and Gas Technology,
November/December 2004.
Appendix
WRC Bulletins
Bibliography of the Pressure Vessel Research Council, Welding Research

Council Bulletins, www.forengineers.org.
494. Fracture-Safe and Fatigue Design Criteria for Detonation-lnduced Loading in
Containment Vessels by Thomas A. Duffey and Edward A. Rodriguez, August 2004,
65 pp. <ISBN: 1-58145-501-1>.
493. Guidelines for Modeling Cylinder-to-Cylinder lntersections by G. E. O. Widera and
Xue Liping, July 2004, 39 pp. (ISBN: 1-58145-500-3).
492. Piping System S!Fs and Flexibility Analysis Criteria by Rudolph J. Scavuzzo and
E. C. Rodabaugh, June 2004, 28 pp. (ISBN: 1-58145-499-6).
491. Interpretive Report on Dynamic Analysis and Thsting of Pressurized Components
and Systems by George Antaki, Jerry Bitner, Keshab Dwivedy, Henry Hwang, and
Rudy Scavuzzo, May 2004, 56 pp. (ISBN: 1-58145-498-8).
490. Damage Mechanisms Af{ecting Fixed Equipment in the Fossil Electric Power
Industry by Jonathan D. Dobis and David N. French, April 2004, 106 pp. <ISBN:
1-58145-497-X).
489. Damage Mechanisms Affecting Fixed Equipment in the Refining lndustry by Jonathan
D. Dobis and Cantwell and Martin Prager, February 2004, 151 pp. (ISBN: 1-58145-
496-1).
488. Dam.age Mechanisms A{fecting Fixed Equipment in the Pulp and Paper lndustry by
Jonathan D. Dobis and David C. Bennett, January 2004, 136 pp. (ISBN: 1-58145-
495-3).
487. PVRC Position on Environmental Effects on Fatrgue Life in LWR Applicatwns by
W. A. Van Der Sluys, December 2003, 59 pp. (ISBN: 1-58145-494-5).
486. lndexing Fracture Toughness Data by W. A. Van Der Sluys, John G. Merkle, Bruce
M. Young, Kenneth K. Yoon, Bryan Hall, Makoto Higuchi, and K. lida, November
2003, 42 pp. <ISBN: 1-58145-493-7).
485. Suitable Heating Conditions in Local Post Weld Heat '[}eatment by Hidakazu
Murakawa, Jianhua Wang Lu Hao, and Yukihiko Horii, October 2003, 43 pp. (ISBN:
1-58145-492-9).
484. Gasketed Joint Emissions and Leakage by O. Sakar, A. Bouzid, M. Derenne, L.
Marchand, and U. Muzzo, August 2003, 38 pp. (ISBN: 1-58145-491-0).
483. Creep Crack Growth. Assessment o{ De{ects in High Temperature Components by
Ashok Saxena and Kee Bong Yoon, July 2003, 119 pp. (ISBN: 1-58145-490-2).
482. Special Finite Elements for Piping Elbows and Bends at High Thmperatures with
Creep by T. L. Anderson and Gregory W. Brown, June 2003, 23 pp. <ISBN: 1-58145-
489-9).
425
426 Appendix WRC Bulletins 427
481. The Eflixt O(Post \\eld Heat Treatment and Notch 1bughness on Welded Jomts and 461. Experimental Leak Te~;ting of 16-inc:h Class 300 RFVt'N Flange with and without
on j'llomwlized Base-Metal Propertítts of A516 Steel by Elmar Upiti~;, Ken Orie, and Externai Bending Morrtcnt by G. Bíbel. T. Fath, W. Palmer, R. Ried~I. and T.
Charles R Ropcr, ~lay 2003. 83 pp. ISBN: 1-58145-488-01. Westlind, May 2001, 54 pp. <ISBN: 1-58145-468--6).
480. Ef{ects o/ Phosphorous and Sulfur on Susceptibility to Vleld Hot CrackinR in 460. High Temperoture Cracking and Properties of Stainless Steel Flux-Cored Welds
Auste111tr. StE·cls by Leijun Lí and Robert W. Messler. April 2003. 26 pp. USBN: and Effects of B1smuth - Repnrt 1: Investigation on High Temperuture Properties of
1-5~145-4~7-2 Weld Metais o{ Stamle.,s Ste..t Flux-Cored Wires and Guidelines by The Japan
479. Simple Formulations to Et•aluate Surface Impacts on Bun<'d Sütel Pípelínes by Welding Engmeeríng Society. Report 2: Po~:ition Statement on the Effect ofBismuth
Abhlna,· Gupta and Rake!<h Kumar Saigal. February 2003. 31 pp. ISB);: 1-58145- on the Elet•ated Tempt!roture Properties of Flux Cored Stainless Steel Weldments
486-4). (with Appendix 011 JIW Round Robin: Bismuth of Stainless Steel FCW Welds) by
478. Stre.~s Intensity and Crack Grmt'lh Opening Area Solutions for Through-Wall Cracks J. C. M. Farrar. A. W. ~larshall, and Z. Zhang, April 2001, 27 pp. ·ISBN: 1-58145-
in Cy/mders and Spheres by Ted L. Ander:;on. January 2003. 31 pp. USBN: 1-58145- 467-81.
485-61. 459. Fracture 1bughness Master Cun•e Det•elopment: Strategies for RPV Assessment by
477. Desígn of Pressure \~ssels for High Stmin Rate Loading: Dynamic Pressure and W. A. VanDerSluys, C. L. Hoffmann. W. L. Server, R. G. Lott, 1\f. T. Kirk, and C. C.
Failure Criteria by Thomas A. Duffey, Edward A. Rodriguez. and Christopher Kim. February 2001, 52 pp. (ISBN: 1-58145-466-XI.
Romero, D(J{'ernbcr 2002, 58 pp. <ISBN: 1-58145-484-81. 458. Fracture 1bughness Jlaster Curve Det•elopment: Application of Master Cun·e Frocture
4 76. Recommcndations for Determining Residual Stresses 111 Fitness-for-Sen·tce Assessment 1bughness Methodology for Ferritic Steels by W. A. Van Der Sluys, C. L. Hoffmann,
by Pingsha Dong and Jeong K Hong, November 2002, 61 pp. IISBN: 1-58145-483-XI. K K Yoon, W. L. Server, R. G. Lott, S. Rosinski, :'.t T. Kirk, S. Byrne, and C. C. Kim,
475. Studies of Local Diff'erences in Material Creep Properties on Weldments by Carl D. January 2001, 36 pp. 'ISBN: 1-58145-465-1).
Lundin. Liu Peng. Ted L. Anderson, and Gregory V. Thorwald, September 2002, 457. Fracture Thughness Master Curl'e Development: Fracture Thughness of Ferritie Steels
135 pp. !ISBN: 1-58145-482-11. and ASTM Reference Thmperature <TAJ by W. A. Van Der Sluys, C. L. Hoffmann, K K
474. Master S-N Curt•e Method for Fatigue Evaluation ofWelded Componmt.~ by Píngsha Yoon. D. E. Killian, and J. B. Hall, December 2000, 56 pp. ISBN: 1-58145-464-31.
Dong, ,Jeong K. Hong, David A. Osage, and Martin Prager, August 2002, 50 pp. 456. Heat Exchanger Flow Charactenzation-HXFLOW Software: Theory Manual and
<ISBN: 1-58145-481-31. Users Manual by D. Mitra-Miijumdar, K K Niyogi, and V. Ratehalli, November 2000.
473. Externai Benclirlf( Jfoments 011 Bolted Gasketed Joints by Yves Birembaut, Hakim 32 pp. !ISBN: 1-58145-463-51
Bouzid, Michel Dcrcnnc, Thicrry Ledauphin, Luc Marchand, Pascal Mart.elli-Garon. 455. Recent Progress m Analysis of Residual Welding Stresses- Report 1: Mod.eling of
and Vincent Masi, July 2002, 56 pp. ISBN: 1-58145-480-51. Weld Residual Stres.->Cs and Dtstortion· Computational Procedures and Applicatwns
472. Using Finitc Elemcnt Anal;ysis for Determining the Bending Moment rB2J P1pi11f( by Pingsha Dong. R~tport 2: Fa.,t Th.-rmal Solution Procedure for Aruúyzing 3D
Elbow Stres., lndir<•s by Vernon C. Matzen and Ying Tan, June 2002. 48 pp. !ISBN: Multi-Pass Welded Structures by Z. Cao, P. Dong, and F. Brust, Report 3: Finite
1-58145-479-1' Element and Experimental Studv of Residual Stresses in a Multi-Pass Repair Weld
471. Dct•elopment of Str.·ss Intensity Factnr Solutions for Surface and Embedded Cracks by P. Dong, J. Zhang..J. K. Hong, W. Bt'll, and E. J. McDonald, September 2000, 28
in APJ 579 by T. A. Andcn:on, Gregory Thorwald, Daniel J. Revelle, David A. Q;;agc. pp. <ISBN: 1-58145-462-71.
Jeremy L..Janelle, and Mntthew E. Fuhry, May 2002. 79 pp.IISB:-;: 1-58145-478-31. 454. A L!teroture Ret·ietL' on Characteristics of High Temperoture Ferritic Cr·Mo Stecl.~
470. Rerommcndations for Dcsign of\éssels for Elet•ated Temperoture Sen;ice by \'incent and Weldments by Carl D. Lundin, Peng Liu, and Yan Cui, August 2000, 36 pp.
A. Carucci, Raymond C. Chao, and Douglas J Stelling, April 2002. 23 pp. CISBN: !ISB~: 1-58145-461-9).
1-58145-477-SI. 453. Minimum W..td Spacing Requiremtnts for API Aboue Ground Storage by J.M. Lieb,
469. Crack-Starter Weld Bead Deposition for ASTM E-208 Drop-V.'eight Testing-Part K Mokhtarian, L.R Shockley, and Elmar Upitis, July 2000, 32 pp. ·ISBN: 1-58145-
1-Qualification of Elcctrodes for the Crack-Starter Based on P-2 Type Drop· Weight 460-01.
Specimen.~--AST.\1 E 201'1-9.Sa-Part 2: Determination of the NDT Temperature 452. Recommended Practiccs for Local Heating of Welds in Pressure Vessels by Joseph
Usinf( Qualífi~:d El,.cfl·odes otl P-2 Type Specimens with Different Strertgth Base W. ~fcEnerney and Ping~ha Dong, June 2000,64 pp. <ISBN: 1-58145-459-7).
Materials-AST.\f E 20b·95a by C. D. Lundin, Peng Liu, Songqum Wen, Ralph 451. Internai Pressure Demgn of l~<olated Nozzles in Cylindri.cal Vessels with d/ D up to
Edwards, and Raymond Bellamy. February 2002, 27 pp. CISBN: 1-58145-476-71. and Includ1ng 1.00- Report 1: Code Rules for Interna/ Pressure Design of Isolated
468. Leak Te,çtinf( of a Raised Face V.eld Neck Flange by G. Bibel, D Weinberger, C. Nozzles in C_vlmdrical Vessels u·ith dI D IH .O by E. C. Rodabaugh, Report 2: Limit
Syverson, and S. Dockter, January 2002, 23 pp. <ISBN: l-58145-475-9J. Analysis and Burst Trst for Large D1ameter Intersections by Z. F. Sang, S. P. Xue,
467. Characterization ofThermo-mechanical Fatigue Response of SS-316 Stmctures by Y. J. Lin and G. E. O Widera. May 2000. 52 pp. IISBN: 1-58145-458-91.
S. Y. Zumr;k, L. C. l<'irth. ~1. L. Renauld. and D. Davis, December 2001, 30 pp. 450. NPS 4 Class 150 Bolted Flanged Joints Subjected to Pressure and Externai Bending
!ISBN: 1-58141)-470-0l. Loads by Luc Marchand. Daniel Laviolette, and Michel Derenne, April 2000, 20 pp.
466. Behavror o( B1•1/ows by C. Bccht, IV, :'ilovember 2001. 80 pp. CISBN: 1-58145-473-2). (ISBN: 1-58145-457-01.
465. Tec:lmolo~ws f'or the Et•aluatwn ofNon-Crack-Like Flaws in Pressurized Compommts 449. Cuide/ines for the Design and lnstallation o{ Pump Piping Systems by Vincent A.
- Ero.~imr I Cnrroswn, Pittmg, Blisters, Shell Out-of-Roundness, Weld Misalignment, Carucci and James C. Payne, February 2000, 48 pp. (!SBN: 1-58145-456-2).
Bulf(<'8. aml Dents, September 2001, 135 pp. <ISBN: 1-58145-472-4). 448. Eualuation of Welded Attaclrments on Ptpe and Elbows by E. C. Rodabaugh, E. A.
464 . Cuide/ines for SiziiiR ofVessel.ç by Limit Analysis by A. Kalnins, August 2001, 16 pp. Wais. and G.B. Rawl!'!, January 2000. 40 pp. <ISBN: 1-58145-455-4).
liSBN: 1-58145-471·6!. 447. Evaluation o{ Operatmg Margins for ltt-seruíce Pressure Equipment by E. Upitis
463. Píp1111: Burst anel Cyclic Moment Testing and Standardized FlexLbilzty Factor Method and Kam Mokhtarian. December 1999, 59 pp. <ISBN: 1-58145-454-6).
by Evcr('tt C. Rodabaugh, E. A. Wais. G. E. Woods, andE. G. Reineke. July 2001, 68 446. Design and Repair of Buried Pipe by George A. AntakJ., November 1999, 38 pp.
pp. <ISBN: 1-58145-470-81. IISBN: 1-58145-453-81.
462. Commentary 011 th• Alternatit·e Rufes for Determining Allowable Cornpres.sive Stn:sses 445. Pla.~tic Pipe: Bur.~l and Fatigue Testirrg of PVC and HDPE Pipe by R. J. Scavuzzo,
for Cylinders, Con., Spheres, and Formed Heads for Sectwtl VIII. Dit·isions 1 a11d 2 M. Cakmak, T. S. SriYatsan, :\f. Cavak, G. E. O. Widera, L. Zhao, H. Chen, P. Hu,
by C. D. Miller, .Jur .. 2001 27 pp.IISBN: 1-58145-469-41. and P. C. Lam, September 1999, 78 pp. HSBN· 1-58145-452-X ..
428 Appendlx WRC Bulletins 429
444. Buckling CritPria for Tori.çhperical Heads under Interna/ Pressure by C. D. Miller, 426. Differential Destgn and ConstructtOII Cosi o(Nuclear Power Plant Piping Systenzs as
August 1999, 99 pp. (ISBN: 1-58145-451-1>. a Function o(Seismic Intensityand Tfme Period o(Construetion by T. M. Adams and
443. Externai Pressure: Effect o( Initial lmperfections and Temperature Limit!l by C. D. J. D. Stevenson, November 1997, 31 pp. <ISBN: 1-58145-433-3).
Miller and D. S. Griffen, July 1999, 40 pp. 1ISBN: 1-58145-450-3). 425. A Ret>iew o( lfethods for the Analysi.~ o( Buried Pressure Piping by G. Antaki,
442. Polytetrafluomethylcne (PTFEI Gasket Qualification by M. Derenne, L. Marchand, September 1997, 29 pp. !ISBN: 1-58145-432-5).
and J. R. Payne, June 1999. 47 pp. 'ISBN: 1-58145-449-Xl. 424. Bibliography of.the Welding Research Supplements o(lhe Welding Journal- Published
441. Devdopmcnl o( a Comprehensil·e Static Seismtc Anal_vs1s Method for Piping in tbe American Weldmg Society l'.'elding Journal from 1950-1996, Compiled by R. La
S_w:tems by Timothy ::'\1. Adams and John D. Stevenson, May 1999, 47 pp. (ISBN: Pointe,August 1997, 190 pp. IISBN:0-9656164-8-7).
1-58145-44ts-ll. 423. Evaluatwn o{ Seismic Response Data for Pipmg by G. C. Slagis, July 1997, 162 pp.
440. A Synthe.sis o( the Fracture Assessment Methods Proposed in the French RCC-.\1R <ISBN: 0-9656164-7-9
Code for High Temperature by D. Moulin. B. Durbay. and L. Laiarinandrasana, 422. Fatigue o( Welded Structure.~ by J. M. Barsom and R. S. Vecchio. June 1997, 64 pp.
April 1999, 108 pp. USBN: 1-58145-447-3>. <ISBN: 0-9656164-6-0
439. Use o( Low Carbon l~CdUfo Weld .\fetal for Fabrication o(Cr·Mo Components by 421. Welding Type 347 Sta nless Stee/ by R. D. Thomas, ~1ay 1997, 127 pp. <ISBN
C. D. Lundin, P. Liu. G. Zhou, and K. Kahn, February 1999, 99 pp. <ISB::-;: 1- 0-9656164-5-21.
58145-446-51. 420. Interpretive Report on Nondestructiuc Examination Thchniques, ProceduT"P For Piping
438. lntermetalltc Phas., Precipitation in Duplex Stainless Steels and Wtld Metais: and Heavy Section Ves>'l'ls by S. H. Bush April 1997, approx. 99 pp. ISBN: 0-9656164-
Metallur10, JnfluenC'e 011 Properties, Welding and Testing Aspects by Lei f Karlsson, 4-4).
January 1999, 23 pp. <ISBN 1-58145-445-7). 419. Eleuated Temperaturo: Characlertzatton o( Flexible Graphite Sheet Materiais for
437. Section 111, Division 3 fNUPACKJ o(The ASME Boiler & Pressure Code: Assessment, Bolted Flanged Joints by M. Derenne. L. Marchand. and J. R. Payne, February
Sample Problems and Commentary on Design by T. M. Adams. December 1998, 42 pp. 1997, 88 pp. CISBN: 0-9656164-3-6).
nSBN: 1-58145-444-9>. 418. Constraint Effects on Fracture Behauior: (1) The Effect o{Crack Depth (a) and Crack·
436. Evaluation o(Small Branch Connections with Thraugh-Run Moments by D. H. Roarty, Depth to Wtdth Ratw (a/Wi on the Fracture Toughness of A533-B Steel, (2) An
E. C. Rodabaugh. E. A. Wais, P Ellenberger, and S. E. Moore, November 1998, 35 pp. AnalyticallnvesfiJ?alion o( the Effect o( Crack Depth (a) and Crack Depth to Width
(ISBN: 1-58145-443-0l. Ratio (a/WJ on the Fracture Toughnes~ o(A533·B Steel, and (3) The Significance o{
435. E~·aluation o(De11ign Margins for Section VIII. Div. 1 and 2 ofthe ASME Boiler and Crack Depth (a) artd Crack Dcpth to Wtdth Ratio (a/W) With Respect to the Behavior
PT"Pssure Vessel Code by E. Upitis and K Mokhtarian September 1998. 85 pp. I ISBN: o(Very Large Specimen.• by J A. Smith and S. T. Rolfe, January 1997, 43 pp. (ISBN:
1-58145-442-21. 0-9656164-2-8).
434. V.•1<ign and ln-.•ervia! Mary;:ins o( Pou·er Piping Systems: A Comparatitoe Study o( 417. Design Guide to Reduce Potentia/ for Vibration Caused by Fluid Flow Inside Pipes-
U.S.. Canadian anel Europcan Codes & Standards by George A Antaki, August. 1998, Review and Survey by C. W. Lin, December 1996. 36 pp. !ISBN: 09656164-1-Xl.
21 pp. !ISBN: 1-58145-441-4\. 416. Creep Crack Growth BehaL•wr in Weld .\fetal I Base Metal I Fusion Zone Regions in
433. Fatigue o( Butt-Weldcd Pipc and Effect o( Testing Methods- Report 1: Fatigue o( Chromium Molybdenum Ste-·ls by R H. Norris andA. Saxena, November 1996, 61
Butt· Welded Pipe b_y R. .J Scavuzzo, T. S. Srivatsan, and P. C. Lam. Repart 2: Effect pp. ISBN: 09656164-0·1).
o(Thsting .\fethods on Strcss Intensification Factors by E. C. Rodabaugh and R. J. 415. Literature Survey and Interpretiue Study on Thermoplastic and Rein{orced·
Scavuzzo, July 1998. 79 pp. (ISBN: 1-58145-440-61. Thermosettzng-Resin Piping and Component Standards by W. E. Short II. G. F. Leon,
432. Fatigue Strpngth Reduction and StT"Pss Concentration Factors for Rélds m Pn·ssuN G. E. O. Widera, and C. G. Ziu. September 1996.28 pp. 'ISBN: 1-58145-414-7).
\~.~se/ and Ptpu~ - Report No. 1: lnterprptitoe Review o( Weld Fatii(Ue·Strcngth· 414. A Neu· Design Criterion Ba ...·d 011 Pre1111ure Testing o( Torispherical Head11 by
Rf'duction anel Stn·ss-Cona!rltration Facton: by C. E. Jaske, Report No. 2: Fatigue· A. Kalnins and M. D. Rana, August 1996. 60 pp. ; ISBN: 1-58145-413-9).
Stn-n!(th·Reduction Factors for Welds Based on ,VDE by J. L. Hechmer andE .•]. Kuhn, 413. Development o( Criteria for Assessment o( Reactor Ve.•sels with Low Upper Shelf
III. June 1998, 55 pp. <ISBN: 1-58145-439-21. Fracture Toughness. July 1996, 53 pp. (ISBN: 1-58145-412-0).
431. Summary o(Ga.~ket.• Steam Leakage Te.çts- Report No. 1: GaBMt Steam Leakage Tests 412. Challenges and Solutions in Repair Welding for Power and Processing Plants -
hy Y Birembaut, T. Lcdauphin, and Y. Morio, &port No. 2: Leak Tests Conducted on Proceedings o( a Workshop, June 1996, 125 pp. 'ISBN: 1-58145-411-2).
Graphito: Gaskets by B. S. Nau and M. D. Reddy, Report No. 3: Long Duration Air and 411. An Experimental Study o( Causes and Repair o( Cracking of 1'/.Cr- !:Mo Steel
Steam Screl'Tling Tests on Ela..•tomeric Sheet Gasket Materiais by L. Man:hand and M. Eqwpment by C. D Lundin, P. Liu, C. Y. P. Qiao. G. Zhou, K K. Khan, and M.
Derenne. May 1998.54 pp. 1ISBN: 1-58145-438-4! Prager, May 1996, 215 pp. (ISBN 1-58145-410-4>.
430. ReL·iew o( Exi!lltnJ( Fitness -For-Servtce Criteria for Crack-Like Flaws by P M. Scott, 410. Evaluation o( Design Criterw for Storage Tanks with Frangible Roof Joints by
T. L. Ander~on, D. A. Osage, and G. M. Wilkowski, April 1998, 155 pp. <ISBN: D. Swenson, D. Fenton, Z. Lu, A. Ghori, and J. Baalman, April 1996, 73 pp. (ISBN:
1-58145-437-6>. 1-58145-409-0l
429. 3D Stress Critena Guideline11 for Application by J. L. Hechmer and G L. Holhnger, 409. Fundamental Studies !1( lhe Metallurgical Causes and Mitigation o( Reheat
February 1998, 137 pp. !ISBN: 1-58145-436-8). Cracking in 1Cr-Mo and 2Cr-1 Mo Steels by C. D. Lundin and K. K. Khan, February
428. Unmixed Zone in Are Welds: Significance on Corrosion Resistance o( High 1996, 117 pp. IISBN: 1-58145-408-2).
Molvbdenum Stainless Steels by C. D. Lundin, W. Liu, G Zhou, and C. Y. Qiao, 408. Bolted Flange Assembly: Preliminary Elastic Interaction Data and Improved Bolt-up
January 1998, 98 pp. (ISBN: 1-58145-435-XJ. Procedures by G. Bibel and R. Ezell, January 1996, 27 pp. CISBN: 1-58145-407-41.
427. Leakage and Emisswn Characteristics o{ Sheet Gaskets: Report No. 1: Fugitit•c 407. Reports on Heat Treatnwnt o( Stel'lH Us••d m Boiler and Pressure Vessel Applications·
Emi.~.~ion Characteristics o( Gaskets and Report No. 2: Exploratory lnt•estigation o( the Cl) Carbon Migration tn Cr-Mo Weldnumts Effect on Metallurgical Structure and
Leakagc S,'abilization Time at Roam Tempera.ture for Flexible Graphite and PTFE Mechanical Properties by C. D. Lundin, K K. Khan, and D. Yang and 21 ASME
Ba:-;ed She- ·t Ga~<kets, M. Derenne. L. ~farchand. and F. Deshaies, December 1997, 61 Post·Weld Heat Treatíng Practiccs.: An Interpretive Report by C. E Spaeder. Jr. and
pp.IISBl\ 1-58145-434-ll. W. D. Doty, December 1995, 65 pp. <ISB:-i: 1-58145-406-6 •.
406. Proposed Rules for D.!termining Allou·able Compresswe Stresses for Cylinders, Cones, and A. Yasuo, and !2) Acoustíc Resonance in Heat Exchanger Thbe Bundles by R. D.
Spheres and Formed Heads by C. D. Miller and K Mokhtarian, November 1995. 24 Blevins, February 1994, 74 pp. 1ISBN: 1-58145-388-4).
pp. CISBN: 1-58145-405-8). 388. Research in the USSR on Residual Stn?sses and Distortion m Welded Structures by
405. Effect of Heat Treatment on the Elevated Temperature Properties of a 2Cr-1Mo V. L Pavlovsky and K Masubuchi, January 1994,64 pp. 'ISBN: 1-58145-387-6).
Submerged Are Weldment by C. D. Lund.in and K. K. Khan. September 1995 45 pp. 387. White Paper on Reactor Vessellntegrity Requirements for Levei A and B Condition.~
ISBN: 1-58145-404-XJ. (Companion Bulletin to 386) by ASME Section XI Task Group on Reactor Vessel
404. Fatigue Cro.<"k Growth of Low-Alloy Steels in Light Water Reactor Em•ironments: (11 Integrity Requirements, December 1993,77 pp. 1ISBN: 1-58145-386-8).
Environmentally-As.~i!<ted Crocking of Ferritic Steels in Aqueous Em•ironments: An 386. lnternational Views on Reactor Pressure Vessel lntegrity rCompanion Bulletin to
lnterpretive Review by L. A. Jame:;. (2J Modeling o{ Fatigue Crock Grou·th Rate for 3871 by ASME Section XI Task Group on Reactor Vessel Integrity Requirements,
Ferritic Steels in Light Water Reactor Environments by E. D. Ea,;on, E. E. Nelson, and November 1993, 44 pp. <ISBN: 1-58145-385-XI
J. D. Gilman, and t3l Technical Basi3 for a Revi.sed Fatigue Crack Growth Rate 385. Joining o{ 6061 Aluminum Matrix-Ceramic Particle Reinforced Composites by R.
Referenc.: Curve for Ferritic Steels in Líght Water Reactor Ent•ironments by E. D. Klehn and T. W. Eagar, September 1993, 26 pp. 1ISBN: 1-58145-384-U.
Ea!;On, E. E. Nelson, and J. D. Gilman. August 1995, 51 pp. <ISBN: 1-58145-403-1). 384. lmproving Steel Spot Weld Fatigue Resistance: Study 1: The Effect of Temper Cycle,
403. Metallurgical and Fracture Toughness Studies o{ A516-70 Steel: (]J Metallur~tical Mechanical Treatments. Weld Geometry and Weldmg Conditions on Sheet Steel Spot
Characterization of the HAZ in A516-70 and Evaluation o{ Fracture Touflhness Weld Fatigue Resistance by F. V. Lawrence, H. T. Corten, and J. C. Mcl\-iahon, and
Specimens by C. F. Lundin, G. Zhou, and K. K Kban, and (li) Comparison o{ the Study li: The Effect of Weld Bonding, Adhesiue Bonding, Weld Metal Expulsion and
CTOD Fracture Toughness ofSimulated and Weldment HAZ Regions 1n A516 Steel Sur{ace Condition on Sheet Steel Spot Weld Fatigue Resi.stance by G. Banas, A.
with Deep and Shallow Cracks, by J A. Smith, R. M. Holcomb, and S. T Rolfe, ,July Cieszkiewicz, and F. V. Lawrence. August 1993, 49 pp. (ISBN: 1-58145-383-3).
1995, 111 pp. <ISBN: 1-58145-402-31 383. Non-Destructive Measurement and Analysis of Residual Stress in arul Around
402. Creep-Fatigue Damage Assessment in Type 316 Stainless Steel under Uniaxial and Welds-A State-of-tlze-Art Survey by IIW Commission V on Quality Control and
Multiaxial Strain Cycling at 1150"F by S.Y. Zamrik and M. Mirdamadi, June 1995, Quality Assurance of Welded Products. July 1993, 17 pp. (ISBN: 1-58145-382-5).
34 pp. (ISBN: 1-58145-401-5). 382. Nucl~ar Piping Criteria for Advanced Light-Water Reactors, Volume 1-Failure
401. Underwater Wct Welding o{Steel, S. lbarra, S. Liu, and D. L. Olson, May 1995, 39 pp. Mechanisms and Corn?ctiue Actions by the PVRC Piping Review Committee, June
<ISBN: I-58145-400-7>. 1993, 45 pp. <ISBN: 1-58145-381-7).
400. lnterpretwe Report o( Weldability Tests for Hydrogen Cracking of Higher Strength 381. Non-ldentical Flanges with Full-Face Elastic Gaskets by A. E. Blach, May 1993, 11
Steels and Tlzeir Potential for Standardization, B. A. Graville, April 1995, 44 pp. pp. <ISBN: 1-58145-380-91.
<ISBN: 1-58145-399-XJ. 380. Recommendations to ASME for Cocle Guidelines and Criteria for Continued Operation
399. Tlze lnfluence ofCon.~umablt> Composition and Solidification on lnclusion Formation o{ Equipment by the PVRC Task Group on Continued Operation of Equipment, April
and Growth in Low Carbon Steel Underwater Wet Welds by A. Sanchez-Osio and S. 1993, 59 pp. <ISBN: 1-58145-379-5).
Liu, February 1995, 59 pp. ,ISBN: 1-58145-398-1). 379. Alternative Methods for Seismie Analysis o{ Piping Systems by the Committee on
398. Reduction o{S-N Curves for Ship Struc:tural Details by K A. Stambaugh, D. H. Le"-"<>n, DynamicAnalysis ofPVRC, February 1993,39 pp. <ISBN: 1-58145-378-7).
F. V. Lawrence, C. Y. Hou. and G. Banas, January 1995, 73 pp. (ISBN: 1-58145-397-3). 378. Review and Evaluation o{the ToughMs.~ ofAu.~t.·nitic Steels and Nickel Alloys A{ter
397. Empirical Modeling for Real-Time l\éld Process Control and Generator Monitoring by Long-Term Elevated Temperatun? Exposures by S. Yukawa, January 1993, 53 pp.
X Xiao~hu, H. Vandcrveldt, and J. Evans, December 1994, 21 pp. CISBN: 1-58145- <ISBN: 1-58145-377-9).
396-5). 377. Development o{ Test Procedures for F1n? Resistance Qualificatíon of Gaskets by M.
396. Research Report on Characteri:zation and Monitoring o{Cracking in \let H2S ç, ,., ice Derenne, J. R. Payne, L. Marchand, andA. Bazel'g\n, December 1992, 19 pp. ISBN
by ~1. S. Cayard, R. D. Kane, L. Kaley, and ~1. Prager, November 1994, 136 pp. 1-58145-376-0 •.
(ISBN: 1-58145-395-7J. 376. Metal Fatigue zn Operating Nuclear Power Plants by ASME Section XI Task Group
395. Vanadium and Columbium Additions in Pressure Vessel Steels by P. Xu, B. R. Somer:;, on Fatigue in Operating Plants, November 1992. 42 pp. r ISBN: 1-58145-375-21.
andA. W. Pense, September 1994. 59 pp. (ISBN: 1-58145-394-91. 375. The Significance of the a/W Ratio on Fractun? Toughness o{ A-36 Steef by R. A.
394. Simplified .\fethods for Creep-FatiguE Damage Evaluations and the Applicatiotl to Whorley and R. T. Rolfe. September 1992, 34 pp. <ISBN: 1-58145-374-4
Li{e Extension by M. J Manjoine, August 1994, 25 pp. <ISBN: 1-58145-393-0l. 374. Papers Presented at the Con{erence on •Life o{ Pressure Vessets• held by the French
393. lnterpretiL-e Report on Dynamic Analysis and Testing of Pressurized Components and AFIAP in 1989, July/August 1992, 91 pp. <ISBN: 1-58145-373-61.
S_vstems-Fifth Edition by J. S. Leung, G. A. Antaki, T. L. Wang, R. D. Blevin~. K. :\1, 373. Research on Modem High Stn?ngth Low Alloy Steel Welding by P. L. Harri~on and
Vashi, and M. S. Whitt, July 1994, 52 pp. <ISBN: 1-58145-392-21. P. H. M. Hart, June 1992, 44 pp. CISBN: 1-58145-372-8).
392. Developing Stress Intensification Factors: (1) Standardized Method for Developing 372. Guidelínes for Flow-Induced Vibration Preuention in Heat Exchangers by J. B.
Stress lntensificatwn Factors for Piping Components by E. C. Rodabaugh and <2) Sandüer, May 1992, 27 pp. (ISBN: 1-58145-371-X).
Effects o{ Weld Metal Profile on the Fatigue Li{e o{ Integrally Rein{orced Weld-on 371. Characterization of' PWHT Behavior of 500 N I mm2 Class TMCP Steels by the
Fittings, by G. E. Woods andE. C. Rodabaugh, June 1994, 26 pp. <ISBN: 1-58145- PVRC Subcommittee on Pressure Vessel Steels, April 1992, 34 pp. <ISBN: 1-58145-
391-4). 370-ll.
391. Elevated Thmperature Testing of Gaskets for Bolted Flanged Connectwns by M. 370. PVRC Committee on Reuiew o( ASME Nuclear Codes and Standards Report of
Derenne, L. Marchand, J R. Payne andA. Bazergui, May 1994, 37 pp. <ISBN: Actiuities, January 1988 to January 1991- Summary ofRecommendations, Presented
1-58145-390-6). by the PVRC Steering Committee, February 1992, 55 pp. <ISBN: 1-58145-369-8).
390. Failurc of Welds at Elevated Temperatures by G. R. Stevick, March 1994, 39 pp. 369. Nitrogen in Are Welding-A ReL·iew, December 1991/January 1992, 171 pp. <ISBN:
CISBN: 1-58145-389-2>. 1-58145-368-Xl.
389 Damping and Resonance o{ Heat Exchanger Thbe Bundles: (1) Vibration Damping o( 368. Stresses in Intersecting Cylinders Subjected to Pressure by K. Moktarian and J. S.
Heat Exchanger Thbe Bundles in Thoo-Pizase Flow by M. J. Pettigrew, C. E. Taylor, Endicott, November 1991, 32 pp. !ISBN· 1-58145-367-1
432 Appendix WRC Bulletlns 433
367. Basis o(Curren t Dynamic Stress Criteria for Pipmg by G. C. Slagis, September 1991, High Strength Low Alloy Steels by R. Menon, C. D. Lundin, and Z. Chen, December
46 pp. CISBN: 1-58145-366 -3). 1989, 30 pp. (ISBN: 1-58145-348-5!.
366. Recommended Practices in Elevated Temperature Design: A Compendium of Breeder 348. Repair Welding of Seruice Exposed Cr-Mo Steel Weldment by C. D. Lundin and Y.
Reactor Experiences (1970-1987)-Volume N- Special Topics by A. K. Dhalla, August Wang. November 1989, 39 pp. <ISBN: 1-58145-347-7).
1991, 105 pp. !ISBN: 1-58145-365-5). 347. Welded Tee Connections ofPipes Exposed to Slowly lncreasing Interna[ Pressure by J.
365. Recommended Practices in Elevated Temperature Desígn: A Compendium of Breeder Schroeder, and (2} Flawed Pipes and Branch Connections Exposed to Pressure Pulses
Reactor Experiences (1970-1987)-Volume lll-ln.elastic Analysis by A. K. Dhalla, July and Shock Waues by J. Schroeder. September 1989, 25 pp. !ISBN: 1-58145-346-9).
1991, 84 pp. <ISBN: 1-58145-364 -71. 346. WFI I PVRC M01n.ent Fatigue Tests on 4x3 ANSI B16.9 Tees by G. E. Woods andE. C.
364. New Design Curves for Torisplzerical Heads, and (2) Elastic-Plastic Analysis of SlzeUs Rodabaugh, August 1989, 8 pp. <ISBN: 1-58145-345-0).
of Reuolution under Axisymmetric Loading by D. P. Updike andA. Kalnins, June 345. Assessing Fracture Toughness and Cracking Susceptibility o{ Sreel Weldments - A
1991, 56 pp. <ISBN: 1-58145-363-9). Reuiew by J. A. Davidson, P. J. Konkol, and J. F. Sovak, July 1989, 43 pp. (ISBN:
363. Recommended Practices in Eleuated Temperature Design: A Compendium of Breeder 1-58145-344-2).
Reactor Experiences (1970-1987)-Volume JJ-Preliminary Design and Simplified 344. Three Dimensional Finite Element Analysis ofPVRC 45 Degree Lateral Model4 (d/D
Methods by A. K Dhalla, May 1991, 86 pp. !ISBN: 1-58145-362-0l. = 0.5. DIT = 40) Under Out-of-Plan.e Moment Loading on Branch Pipes. and (2)
362. Recom1nended Practices in Eleuated Temperature Design: A Compendium of Breeder Three Dimensional Finite Element Analysis of 45 Degree Lateral Model2 (d/ D - 0.5,
Reactor Experiences 11970·1987) Volume 1- Current Status and Future Directions by DIT- 10) Under Out-of-Plane Moment Loading on the Branch Pipe by P. P. Rl\iu,
A. K. Dhalla. April 1991, 101 pp. CISBN: 1-58145-361-21. June 1989, 17 pp. CISBN: 1-58145-343-4>.
361. Improuements on Fatigue Analysis Methods for the Design of Nuclear Components 343. Destructiue Examination o{ PVRC Specimen 202 Weld Flaws by Y. Saiga, (2)
Subjected lo the French RCC-M Code by J. M. Grandemange, J. Heliot, J. Vagner, Destructive Examination o{PVRC Nozzk Weld Specimen 203 Weld Flaws by Y. Saiga,
A. ~1orel and C. Faidy and I 21 Framatome Vzew on the Comparison Between Class and <3J Destructiue Examination o(PVRC Specün.en 251J Weld Flaws by S. Ynkawa,
1 and Class 2 RCC-M Piping Design Rules by C. Heng and J. M. Grandemange, May 1989, 47 pp. CISBN: 1-58145-342-6).
February 1991, 26 pp. (ISBN: 1-58145-360-4). 342. Stainless Steel Weld Metal: Prediction of Ferrire Conrent by C. N. McCowan, T. A.
360. Stress Indices, Pressure Design and Stress lntensification Factors for Laterais in Siewert, and D. L. Olson, April 1989, 36 pp. <ISBN: 1-58145-341-8).
Piping by E. C. Rodabaugh, January 1991, 15 pp. <ISBN: 1-58145-359-0). 341. A Preliminary Eualuation of tlze Eleuated Temperature Behavior of a Bolred Flanged
369. Weldability ofLow-Carbon Micro-Alloyed Steels for Marin.e Structures by C. D. Lundin, Conn.ection by J. H. Bickford, K. Hayashi, A. T. Chang, and J . R. Winter, February
T. P. S. Gill, C. Y. P. Qiao, Y. Wang, and K. K Khan, December 1990, 103 pp. (ISBN: 1989, 24 pp. IISBN: 1-58145-340-X).
1-58145-358-2). 340. Interpretiue Report on tlze Mechanical Properties ofBrazed Joints by M. M. Schwartz,
358. Tlze Effect o(Crack Depth to Specün.en Width Ratio on the Elastic-Plastic Fracture January 1989. 55 pp. (ISBN: 1-58145-339-ôl.
Toughness of a Higlz-Strength Low-Strain Hardening Steel by J. A. Smith and S. T. 339. Deuelopment of'Ifghtness Test Procedures for Gaskets in Eleuated Temperature Seruice
Rolfe, November 1990, 19 pp. CISBN: 1-58145-357-4). by A. Bazergui and L. Marchand, December 1988,20 pp. <ISBN: 1-58145-338-8).
357. Calculation of Electrical and Thermal Conductiuities of Metallurgical Plasmas by 338. Interpretiue Report on Electroslag, Electrogas, and Relared Welding Processes by R.
G. J. Dunn and T. W. Eagar, September 1990, 21 pp. IISBN: 1-58145-356-6). D. Thomas, Jr. and S. Liu, November 1988, 29 pp. (ISBN: 1-58145-337-XJ.
356. Finite Element Modelling of a Single Pass Weld by C. K. Leung, R. J. Pick, and D. 337. Experi~n.ental Validation of the Eualuation of Reinforced Openings in Large Steel
H. B. Mok, (2 J Finite Element Analysis of Multi-Pass Welds by C. K. Leung and R. Pressure Vessels by J. Schroeder, October 1988, 22 pp. (ISBN: 1-58145-336-1).
J. Pick, and <3) Thermal and Mechanical Simulations o{ Resistance Spot Welding 336. Interpretiue Report on Dynamic Analysis of Pressure Components - Fourth Edition,
by S. D. Sheppard. August 1990, 41 pp. (ISBN: 1-58145-355-8). September 1988, 30 pp. <ISBN: 1-58145-335-3).
355. Programming and Control o(Welding Processes-Experience ofthe USSR by V. Malin, 335. A Revzew of Area Replacement Rufes for Pipe Connections in Pressure Vessels and
July 1990, 43 pp. CISBN: 1-58145-354-Xl. Piping by E. C. Rodabaugh, August 1988, 57 pp. <ISBN: 1-58145-334-5).
354. Failure Analysis ofa Service-Exposed Hot Reheat Steam Line in a Utility Steam Plant 334. Reuiew o{ Properties of Tlzermo-Mechanically Controlled Processed Steels - Pressure
by C. D. Lundin, K K. Khan, D. Yang, S. Hilton, and W. Zielke and (2) Tlze lnfluence Vessel Steels for Low-Temperature Seruice, June 1988, 49 pp. <ISBN: 1-58145-333-7).
o{ Flux Composition ofthe Eleuated Temperature Properties of Cr-Mo Submerged Are 333. Bibliography on Fatigue of Weldments and Lirerature Reuiew on Fatigue Crack
Weldments by J. F. Henry, F. V. Ellis. and C. D. Lundin, June 1990. 132 pp. <ISBN: Initiation {rom Weld Discontinuities by C. D. Lundin, May 1988, 34 pp. (ISBN·
1-58145-353-lJ. 1-58145-332-9).
353. Position Paper on Nuclear Plant Pipe Supports by Task Group on Nuclear Plant Pipe 332. Characteristics of Heavyweight Wide-Flange Structural Shapes by J. M. Barsom
Supports, May 1990, 51 pp. <ISBN: 1-58145-352-3). and B. G. Reisdorf; and (2) Data Suruey on Mechanical Property Characterization of
352. Independent Support Moti011 ([SMJ Method ofModal Spectra Seismic Analysis by the A588 Sreel Plates and Weldments by A. W. Pense,April1988, 35 pp. (ISBN: 1-58145-
Technical Committee on Piping System Task Group on ISM of the Pressure Vessel 331-0l
Research Council, April 1990, 43 pp. <ISBN: 1-58145-351-5). 331. MetaUurgical Investigation on tlze Scatter of 1bughness in the Weldment of Pressure
351. An Analytical Comparison o(Short Crack and Deep Crack CTOD Fracture Specimens Vessel Steels- PartI: Current Cooperatiue Research, and (2l Metallurgicallnuestigation
o{ An A36 Steel, (2J The Effects of Crack Depth on Elastic-Plasti.c CTOD Fracture on tlze Scatter of Tbuglzness in the Weldment of Pressure Vessel Steels - Part 11:
Touglzness, and (3JA Comparison oftlze J -Integral and CTOD Paramerers for Short Cooperatiue Research, February 1988, 67 pp. CISBN: 1-58145-330-2).
Crack Specimen Testing by W.A. Sorem, R. H. Dodds, Jr. and S. T. Rolfe, February 330. Tlze Fracture Behauior of A588 Grade A and A572 Grade 50 'P.'eldments by C. V.
1990, 34 pp. <ISBN: 1-58145-350-7). Robino, R. Varughese, A. W. Pense, and R. C. Dias, and (2) Effects of Long-Time
350. Design Criteria for Dissimilar Metal Welds by R. H . Ryder and C. F. Dahms, January Postweld Heat 'Ireat~n.ent on the Properties o(Constructional-Steel Weldments by P. J.
1990, 11 pp. (JSBN: 1-58145-349-3>. Konkol, January 1988, 26 pp. <ISBN: 1-58145-329-9).
349. Postweld Heat Treatment Cracking in Chromium-Molybdenum Steels by C. D. Lundin, 329. Accurocy o{ Stress Intensifi.ca-tion Factors for Branch Conn.ections by E. C. Rod.abaugh,
J. A. Henning. R. Menon, and J. A Thdd and (2) Postweld Heat 'Ireatment Cracking in December 1987.44 pp. <ISBN: 1-58145-328-0l.
328. Specimen Thickness Effects for Elastic-Plastic CTOD Toughness o{ an A36 Steel by G. 310. Damage Studies in Prcssure Vessel Companents by F. A. Leckie, December 1985, 30
W. Wellman, W. A. Sorem, R. H. Dodds, Jr., and S. T. Rolfe, <21An Analytical and pp. CISBN: 1-58145-309-4).
Experimental Comparison of Rectangular and Squnre CTOD Fracturc Specimens o{ 309. Deueloprne11t ofa Production Test Procedure for Gaskets by A. Bazergui, L. ~1archand,
011 A36 Stn·l by W. A. Sorem, R. H. Dodds, Jr.. and S. T. Rolfe, November 1987, 23 pp. and H. D. Raut, November 1985, 39 pp. IISBN: 1-58145-308-6).
(ISBN: 1-58145 -327-21. 308. Verification and Application of an /nelastic Analysis Method for LMFBR Piping
327. Long-RaTII-!e Plan for Pressure \<íossel Research · Eighth Editüm by the Pressure Vessel System by H. D. Hibbít and E. K. LeUJ g. September 1985. 28 pp. (ISBN: 1-58145-
Re.-;t.>arch Committee. October 1987, 41 pp. «<SB~: 1-58145-326-4). 307-8).
326. Revised Bul/etin 191. Subrnerged Are-'Welding Procedures for Steels Meeting StaiUÚJrd 307. Fatigue and Creep Rupture Damage ofPerforated Plotes Subjected to Cyclic Plastic
Specifications by C. W. Ott and D. J. Snyder 1this revision is a part o f the Fourth Straining in Creep Regime by M. L. Badlani. T. Tanaka. J. S. Porowski and W. J.
Edition of the WRC Book Weldability o{ Steels. and can only be obtained by ordering O'Donnell, August 1985, 10 pp. <ISBX: 1-58145-306-X •.
the bookl. August 1987, 9 pp.IISBX 1-58145-325-6). 306. PVRC Flanged Joint User Sun'f!_y by J. R. Payne. July 1985, 39 pp. ISBN: 1-58145-
325. Further Gasket Leakal{e Behavior Trends by A. Bazergui, L. Marchand, and H. D. 305-1).
Raut, July 1987, 10 pp. (ISBN: 1-58145-324-81. 305. Summary Reports Prepared by the JPVRC Subcommittee 011 Hydrogen
324. lntoestigation of Design Criteria for Dynamic Loads on Nuclear Power Piping by R .J. Embnttlement: ll) Hydrogen Attack Limit of2'/.Cr-1 Mo Steel by Task Group I, (2)
Scavuzzo and P. C. Lam, June 1987. 20 pp. IISBN: 1-58145-323-XI. Embrittlernent of Pressure Vessel Steels in High Temperature, High Pressure
323. Monograph on Narrow-Gap Welding Technology by V. Malin, May 1987,52 pp. !ISBN: Hydrogen Ent>1ronment by Task Group II. and (3) Hydrogen Embrittlement ofBond
1-58145-322-11. Structure Between Stainless Steel Ouerlay and Base Metal by Task Group lll, June
322. Strain Aging Behauior of Microalloyed Steels by W. A. Herman, M. A Erazo, L. R. 1985, 39 pp. CISBN: 1-58145-304-31.
DePatto, M Sekizawa. andA. W. Pense, !2) The Fracture Toughness Behau10r o( 304. Experimental Limit Coupl~s for Branch Moment Loads on 4-in. ANSI 816.9 TEES
ASTM A737 Grade B and Grade C Microalloyed Pressure Vessel Steels by J. A by J. Schroeder, May 1985, 32 pp. <ISBN: 1-58145-303-5}.
Aadland, J. I. Qureshi, andA W. Pense, (3) The Fracture Behauior of ASTM A737 303. Interpretiue Report on Dynamic Analysis o{Pressure Compone11ts- Third Ed.ition, April
Grade B and Grade C Microalloyed Steel Weldments by J. M. Aurrecoechea, Bi-Nan 1985,33 pp. <ISBN: 1-58145-302-7).
Qain, and A W Pense, and (4) Long Time Stress Relief Effects in ASTM A737 302. Postweld Heat Treatment o{ Pressure Vessel by R. D. Stout, Relaxation Stresses in
Grade B and Gradl' C Microalloyed Steels by N. Shinohe, M. Sekizawa, and A W. Pressure Vessels by P S. Chen. W. A. Herman, andA W. Pense,AStudy o{Residual
Pense, April 1987, 41 pp. IISBN: 1-58145-321-3}. Stress in Pressure Vessel Steels by R. J. Zhou, A. W. Pense, M. L. Basehore, and D.
321. The Dynam1c Deformatio" of Piping by J. L. McLean, P. K. Beazley, and A H H. Lyons, February 1985, 32 pp. <ISBN: 1-58145-301-91.
Manhardt, January 1987, 50 pp. <ISBN: 1-58145-320-51. 301. A Parametric Three-Dimensional Firlite Element Stud.v of 45 Degree Lateral
320. ~Velding .Uetallurgy and Weldability ofHigh-Strength Aluminum Alloy.~ by S . Kou, Connections by P. P. Raju, January 1985, 33 pp. (ISBN: 1-58145-300-0
December 1986. 20 pp. IISBN: 1-58145-319-11. 300. Thchnical Position on Criterza E~>tablishment. 121 Thchnical Position on Dampzng
319 Sensitization of Au.,tenitic Stainless Steels, Effect o{ Welding Variables on HAZ Values for Piping-lntenm Summary Rc>port, !3J Technical Position on Response Spectra
S••nsitization of AIS/ 304 and HAZ Behal•ior of BWR Alternatíve Alloys 316 VG Broadening, and (41 Technical Posítion 011 lndu.~try Practice, December 1984. 38 pp.
c·- d .147 by C. D. Lundin. C. H. Lee. R. ~lenon, and E. E. Stansbury, Nove"l!x>r !ISBN: 1-58145-299-3J.
1986, 75 pp. r ISBN: 1-58145-318-31. 299. Engin.eering Aspects of CTOD Fractun: Toughne.~ Testing by G. W. Wellman and S.
318. Factors lnflul'nci~ the .\feasurernent o{ Ferríte Content in Austenitic Sta11zlcss T. Rolfe, 12< Three-Dimensional Elastic-Piastic Finite Elernent Analysis of Three-
Steel Weld Metal Using Magnetic Instruments by E. W. Pickering, E. S. Robitz. and Point Bend Specimm by G. W. Wellman, S. T. Rolfe and R H. Dodds, and (3! Failu-e
D. J\1. Vandergriff, (2) Measurement of Ferrite Content in Auste11itic Stamless Stcel Prediction ofNotched Pressure \<ío.,,<;e[.'< Using the CTOD Approach by G. W Wellman,
Weld Metal Giving lnternationally Reproducible Results by E. Stalmn:<ek, S. T. Rolfe. and R. H. Dodds, November 1984. 35 pp. !ISBN: 1-58145-298-5).
September 1986, 98 pp. !ISB~: 1-58145-31í-5). 298. Long-Range Plan for Pressurc Ves.•el RPsearch · Seuenth Edition by Pressure Vessel
317. P"-'RC Centri(llf(al Pump-Pipíng Interaction Experience Sun•ey by J. R. Payne. Research Commíttee, September 1984. 39 pp. 'ISBN: 1-58145-297-71.
August 1986, 76 pp. ISBN: 1-58145-316-71.
r 297. Local Stresses in Cylindrical Sh.·lls Dueto Externa/ Loadings on Nozzles. Supplernent
316. Technical Position on Piping System lnstallation Tolerances by E. B. Branch, N. to lVRC Bulletin 107 · !Reuision JJ by J L. Mershon, K :Mokhtarian, G V. Ranjan,
Kalyanam. D. F. Landers. E. O Swain, and D. A. Van Duyne, and (21 Technü:al andE. C. Rodabaugh, August 1984, revised September 1987, 88 pp. !ISBN: 1-58145-
Position or1 Dampíng Value:< for lnsulated Pipe- Summary Report by J. L. Bitner, 296-91.
S. N Hou, W. J. Kagay, and J. A. O'Brien. July 1986, 18 pp. ISBN: 1-58145-315- 296. Fitness-for-Seroice Criteria for Pipelwe Girth-Weld Quality by R. P. Reed, M. B.
91 Kasen, H. I. McHcnry, C. M. Fortunko, and D. T. Read. July 1984, 80 pp. CISBN·
315. Stress Rupture Behauior of Postweld Heat TI-eated 2',;Cr-1Mo Steel Weld Metal by 1-58145-295-0).
C. D. Lundin, S. (' Kelley, R Menon, and B. J. Kruse, June 1986, 66 pp. !ISBN 1 295. Fundamentais o{Weld Discontinuities and Thezr S1gni{icance by C. D. Lundin, June
58145-314-01. 1984, 33 pp. CISBN: 1-58145-294-2).
314. Bolted Flanged Connectiorzs with Fu/1 Face Gaskets by A. E. Blach, A. Bazergui, 294. Creep ofBolted Flanged Connect10ns by H. Krause and W. Rosenkrans, and (2) Short
and R. Baldur, May 1986, 13 pp. <ISBN: 1-58145-313-21. Term Creep and Relaxation Behauior of Gaskets by A Bazergui, May 1984, 22 pp
313. Computer Program.~ for lmperfection Sensitiuity Analysis of Stiffened Cylindrical <ISBN: 1-58145-293-41.
Shell.~ by R. L. Citerley,April1986, 17 pp. (ISBN: 1-58145-312-41. 293. Current Welding Research Problems compiled and edited by R. A Kelsey, G. W.
312. Joining of Molybderwm Base Metais and Factors Which l11fluem:·e Ductibty by A. ,J. Oyler, and C. R. Felmley, Jr., April 1984, 64 pp. HSBN: 1-58145-292-6).
Bryhan. February 1986. 21 pp. <ISBN: 1-58145-311-61. 292. PVRC Milestone Gasket Test., - First Results by A Bazergui and L. Marchand,
311. Assessment ofthe Significance ofWeld Discontinuities. Effect.ç of Microstructure and February 1984, 36 pp. 1ISBN 1-58145-291-8!.
Di..,continuitie., Upon Fracture Morphologv by C. D. Lundin and C. R. Patriarca, 291. Fracture Control ofPres.'<ure \t.ossc/.~ up to 2 Inches Thick by P O. Metz, January 1984,
January 1986, 40 pp. (ISBN: 1-58145-310-8 . 17 pp. !ISBN: 1-58145-290-Xl.
290. Factors Affecting Porosity in Aluminum Welds -A Reuiew by J. H. Devletian and W. 268. Reuiew ofWorldwide Weld Discontinuity Acceptance Standards by C. D. Lundin, June
E. Wood, December 1983, 18 pp. (ISBN: 1-58145-289-6). 1981, 25 pp. (ISBN: 1-58145-267-5).
289. Hot Cracking Susceptibility ofAusten.itic Stainless Steel Weld Metais by C. D. Lundin 267. Elastic-Plastic Buckling of Internally Pressurized Ellipsoidal Pressure Vessel
and C.P. D. Chou, November 1983, 86 pp. (ISBN: 1-58145-288-8). Heads by D. Bushnell, May 1981, 28 pp. (ISBN: 1-58145-266-7).
288. Fracture of Pipelines and Cylinders Containing a Circumferential Crack by F. 266. Weldability and Fracture 1bughness o{ 5% Nl Steel- Part 1: Weld Simulation Testing,
Erdogan and H. Ezzat, October 1983, 23 pp. (ISBN: 1-58145-287-X). and (2) Weldability and Fracture 1bughness of 5% Nl Steel - Part 2: Wuie Plate Thsting
287. Welding ofCopper and Copper-Base Alloys by R. J. C. Dawson, September 1983, 17 by A. D. Dhooge, W. Provost, and A. Vmckier, April 1991, 19 pp. (ISBN: 1-58145-
pp. (ISBN: 1-58145-286-1). 265-9).
286. Fatigue Behavior ofAluminum Alloy Weldments by W. W. Sanders, J r. and R. H. Day, 265. ln.terpretiue Report on Small Scale Thst Correlations with KIC Data by R. Roberts
August 1983, 21 pp. (ISBN: 1-58145-285-3). and C. Newton, February 1981, 18 pp. (ISBN: 1-58145-264-0).
285. Stress Indices and Flexibility Factors for Concentric Reducers by E. C. Rodabaugh and 264. The lnfluence o{ Multiaxial Stress on Low-Cycle Fatigue of Cr-Mo-V Steel at 1000°F
S. E. Moore, and (2) Finite Elemen.t Analysis of Eccentric Reducers and Comparisons by R. H. Marloff and R. L. Johnson, December 1980, 21 pp. (ISBN: 1-58145-263-2).
with Concentric Reducers by R. R. Avent, M. H. Sadd, and E. C. Rodabaugh, July 263. An Annotated Bibliography on the Significance, Origin and Nature ofDiscontinuities
1983, 48 pp. <ISBN: 1-58145-284-5). in Welds, 1975-1980 by C. D. Lundin and S. J. Pawel, November 1980, 14 pp. <ISBN:
284. The Extemal Pressure Collapse Thsts ofThbes by E. Tscboepe and J. R. ~1aison, April 1-58145-262-4).
1983, 20 pp. <ISBN: 1-58145-283-7). 262. Derivation ofASME Code Formulas for the Design ofReverse Flanges by E. O. Waters
283. A Criticai Evaluation ofFatigue Crack Growth Measurement Techniques for Eleuated and R W. Scbneider, (2) Functional Test of a "kssel with Compact Flanges in Metal-
Thmperature Applications by A. E. Carden, February 1983, 11 pp. (ISBN: 1-58145- to·Metal Contact by J. Webjorn and R. W. Schneider and (3) lnterpretiue Report on
282-9). Gasket Leakage Thsting by H. Kraus, October 1980, 33 pp. (ISBN: 1-58145-261-6).
282. Elastic-Plastic Buckling o{ Axially Compressed Ring Stiffened Cylinder-Test us. 261. Effects o{Porosity on the Fracture 1bughness of5083, 5456, and 6061 Aluminum Alloy
Theory by D. Bushnell, November 1982, 28 pp. (ISBN: 1-581451-281-0). Weldments by W. A. McCarthy, Jr., H. Lamba, and F. V. Lawrence, Jr., September
281. Hydrodynamic Response of Fluid Coupled Cylinders: Simplified Damping and 1980, 14 pp. (ISBN: 1-58145-260-8).
Inertia Coefficients by S. J. Brown, October 1982, 37 pp. (ISBN: 1-581451-280-2). 260. Energy Dissipation Characteristics of Pipes and Short Compression Members as
280. The Varestraint Test by C. D. Lundin, A. C. Lingerfelter, G. E. Grotke, G. G. Elements ofPipe-Whip Restraint by S. S. Palusamy, R. L. Cloud, and T. E. Campbell,
Lessmann, and S. J. Matthews, August 1982, 19 pp. <ISBN: 1-58145-279-9). August 1980, 10 pp. (ISBN: 1-58145-259-4).
279. Weldability and Fracture Toughness of Quenched and Tempered 9% Nickel Steel: 259. Analysis o{ the Radiographic Eualuation of PVRC Weld Specimens 155, 202, 203,
PartI - Weld Simulation Testing, and Part li - Wide Plate Testing by A. Dhooge, W. and 251J by E. H. Ruecber and H. C. Graber, June 1980, 35 pp. (ISBN: 1-58145-
Provost, andA. Vmckler, July 1982, 18 pp. (ISBN: 1-58145-278-0). 258-6).
278. The Crack Arrest Properties of9% Nickel Steels for Cryogenic Application.s by R. D. 258. lnternational Benchmark Project on Simplifwd Methods for Eleuated Temperature
Stout and A.W. Pense, June 1982, 20 pp. (ISBN: 1-58145-277-2). Design and Analysis: Problem I -The Oak Ridge Pipe Ratchetting Experiment,
277. High Temperature Properties o{2Y.Cr-1Mo Weld-Metal by C. D. Lundin, B. J. Kruse, Problem li - The Saclay Fluctuating Sodium Levei Experiment by H. Kraus, May
and M. R. Pendley, May 1982, 27 pp. !ISBN: 1-58145-276-4). 1980, 20 pp. <ISBN: 1-58145-257-8).
276. A Summary and Criticai Eualuation of Stress ln.tensity Factor Solutions of Comer 257. Analysis of the Ultrasonic Examinations of PVRC Weld Specimens 155, 202, and
Cracks at the Edge of a Hole by R. L. Cloud and S. S. Palusamy, April 1982, 30 pp. 203 by Stan.dard and Two-Point Coincidence Methods by R. A. Buchanan and O. F.
(ISBN: 1-58145-275-6). Hedden, February 1980, 40 PP- (ISBN: 1-58145 -256 -X).
275. The Use o{Quenched and Thmpered 2'!.Cr-1Mo Steel for Thick Wall Reactor "kssels 256. Review of Data Releuant to the Design ofThbular Joints for Use in Fixed Offshore
in Petroleum Refinery Processes: An Interpretiue Reuiew of25 ~ars ofResearch and Platform.s by E. C. Rodabaugh, January 1980, 83 pp. (ISBN: 1-581452-255-1).
Applicatian by W. E. Erwin and J. G. Kerr, February 1982, 63 pp. (ISBN: 1-58145- 255. Experimental Investigation of Commercially Fabricated 2:1 Ellipsoidal Heads
274-8). Subjected to Internai Pressure by Special Task Group o{ the PVRC Subcommittee
274. lntemational Benchmark Project on. Simplified Methods for Eleuated Temperature on Shells, December 1979, 53 pp. (ISBN: 1-58145-254-3).
Design and Analysis: Problem li- The Saclay Fluctuating Sodium Levei Experiment, 254. A Critical Evaluation o{ Plastic Behauior Data and a Unified Definition of Plastu:
Comparison of Analytical and Experimental Results, Problem III-The Oak Ridge Loads for Pressure Components by J. C. Gerdeen, (2) lnterpretiue Report on Limit
Nozzle to Sphere Attachment by H. Kraus, January 1982, 16 pp. (ISBN: 1-58145- Analysis and Plastic Behauior of Piping Products by E. C. Rodabaugh, and (3)
273-X). Interpretive Report on Limit Analysis o{ Flat Circular Plates by W. J. O'Donnell,
273. Design I mplica.tions ofRecent Aduances in Eleuated Temperature Bounding Thchniques November 1979, 90 pp. (ISBN: 1-58145-253-5).
by J. S. Porowski, W. J. O'Donnell. and M. Badiani, December 1981, 12 pp. (ISBN: 253. A Suruey of Simplified Inelastic Analysis Methods by R. E. Nickell. October 1979.
1-58145-272-1). 18 pp. (ISBN: 1-58145-252-7).
272. Design ofBeam Columns with Lateral-Torsional End Restraint by T. L. Hsu and G. 252. Ultrasonic Eualuation and Sectioning of PVRC Plate Weld Specimen 201 by A. C.
C. Lee and (2) Tapered Columns with Unequal Flanged by G. C. Lee and T. L. Hsu, Adamonis andE. T. Hughes, September 1979, 39 pp. (ISBN: 1-58145-251-9).
November 1981, 23 pp. (ISBN: 1-58145-271-3). 251. Comparison o{ Three-Dimensional Finite Element and Photoelastic Results for
271. Methods o{Analysis of Bolted Flanged Connections- A Reuiew by A. E. Blach and Lateral Connection, WC-12B2, August 1979, 18 pp. (ISBN: 1-58145-250-0).
A. Bazergui, (2) Gasket Leakage Behavior Trends by H. D. Raut, A. Bazergui, and 250. Generalized Y'wld Surfaces for Plates and Shells by D. B. Peterson, W. C. Kroenke,
L. Marchand, October 1981, 42 pp. (ISBN: 1-58145-270-5). W. F. Stokey, and W. J. O'Donnell, July 1979, 18 pp. (ISBN: 1-58145-249-7).
270. Long-Range Plan for Pressure Vessel Research - Sixth Edition September 1981 249. Reuiew ofAnalytical and Experimental Thchniques for lmprouing Structural Dynamic
36 pp. USBN: 1-58145-269-1). ' ' Models, by P. lbanez, June 1979, 44 pp. (ISBN: 1-58145-248-9).
269. lnterpretiue Report on Dynamic Analysis ofPressure Companents- Second Edition, 248. Allowable Axial Stress of Restrained Multi-Segmen.t, Tapered Roof Girders by G. C.
August 1981, 26 pp. CISBN: 1-58145-268-3). Lee, Y. C. Chen, and T. L. Hsu, May 1979, 28 pp. (ISBN: 1-58145-247-0).
247. Corrosion Resistance o{ Brazed Joints by Nancy C. Cole, April 1979, 41 pp. IISBN: 224. lnterpretive Report on Underwater Welding by C.·L. Tsai and K Masubuchi, February
1-58145-246-2). 1977, 37 pp. (ISBN: 1-58145-223-3).
246. lnterpretit·e Report on Dynamic Analysis of Pressure Components. February 1979, 223. Hot Wire Welding and Surfacing Thchmques by A. F. Manz, January 1977, 19 pp.
19 pp. IISB~: 1-58145-245-4). <ISB~: 1-58145-222-5>.
245. A Fracturr Mechanics Evaluation of Flaws in Pipeline Girth Welds by R. P. Reed, 222. The Significance of Weld Discontinuities-A Review o{ Current Literature by C. D.
H. I. McHenry, and l\1. B. Kasen, January 1979. 23 pp. <ISBN: 1-58145-244-6). Lundin, December 1976, 32 pp. 1ISBN: 1-58145-221-7).
244. Wtrasomc Soldering by M. ~~- Schwartz, and 12) Brazir1g 111 a Vacuum by ~1. M. 221. Analysis of Test Data on PVRC Sp.·cífication No. 3, Wtrasonic Examination of
Schwartz. December 1978, 37 pp. 1ISBN: 1-58145-243-81. Forgings, Rensúms I and 11 by R. A. Buchanan. and (21 Analysis of the
243. Effective Utillzation ofHil(h Yield Strength Steels in Fatigue by R. A. l.\Iay, S. Stuber, Nondestructit·e Exammation o( PVRC Plate-We/d Specimen 251J-Part A by R. A.
and S. T. Rolfe, and (2) lnfluence of Yield Strength on Anodic Stress Corrosion Buchanan, November 1976, 29 pp. (ISB~: 1-58145-220-9).
Cracking Resistanee of We/dable Carbon and Low Alloy Steels u·ith Yield Strengths 220. Metal-to-Metal Adhesil·e Bonding by J. P. McNally and C. R Ronan, October 1976.
Below 100 KSI by R. S. Tre:;eder, November 1978. 34 pp. <ISBN: 1·58145-242-XI. 22 pp. 1ISBN: 1-58145-219-5).
242. Fatigue Behat•ior of5000 Series Aluminum Alloy Weldments m Marine Ent•ironment 219. Experimental lm·estigatwn of Limit Loads of Nozzles in Cylindrical \kssels by F.
by W. W, .Sanders, Jr and K. A. McDowell, October 1978, 14 pp. <ISBN: 1-58145- Ellyin. September 1976, 14 pp. (ISBN: 1-58145-218-7).
241-11. 218. Tests o[ Bolted Beam·to·Column Flange .\foment ConnectwllS by K F. Standig, G.
241. Long-Range Plan for Pre:;sure Vessel Research - Fifth Edition by Pressure Vesscl P. Rentschler, and W. F. Chen. August 1976, 17 pp. IISB~: 1-58145-217-9 •.
Re,-earch Committt>c, September 1978, 36 pp. <ISBN: 1-58145-240-3). 217. Through Thtckness Fatigue Properties o[ Steel Plate by C. J. Adams andE. P. Popov,
240. Hydrogcn Embrittlement of Austenitic Stainless Steel Weld Metal with Spec1al and 12) Properties o{ Heavy Section Nuclear Steels by J. M. Hodge, July 1976, 22
Consid!'ration Git•en to the Effects o{Sigma Phase by E. W. Johnson, and S .•]. Hudak, pp. OSBN: 1-58145-216-0l.
August 1978, 51 pp. cISBN: 1-58145-239-XI. 216. Preventmg Hydrogen-lnduced Cracking After Vo'elding of Pressure \kssel Steels by
239. Ret•icw of Fracture :\1echamcs Approaches to Defining Criticai Size Girth Weld Use of Low Temperature Postwcld Heat Treatments by J. S. Caplan and E.
Discontúwitws by G. M Wilkowski and R. J. Eider, July 1978, 25 pp. IISBN: Landerman, .June 1976, 23 pp. I ISBN: 1-58145-215-2).
1-58145-238-ll. 215. Deuelopment of Design Rufes for Dished Pressure Vessel Heads by E. P. Esztergar,
238. Plastic Stability of Pipes and Tee.~ Exposed to Externai Couples by J. Schrocder nnd and (2) The Effect o[Geometrica/ Variations on the Limit Pressures for 2:1 Ellipsoidal
P. Tugcu. June 1978, 23 pp. ciSBN: 1-58145-237-31. Head \kssels Under lntemal Pressure by J. C. Gerdeen. May 1976, 41 pp. (ISBN·
237. lnvestigation of .lft>tlwds of Controlling and Reducing Weld Distortlon in 1-58145-214-4).
Aluminum Structures by K. Masubuchi, May 1978, 28 pp. IISBN: 1-58145-236-51. 214. Stud Welding by T. E. Shoup, April 1976. 22 pp. •ISBN: 1-58145-213-6).
236. Determination of Stiffness and Loading in Bolted Joints Having Circular Geomt>try 213. Weldability of Niobium Containing High Strength Low Alloy Steel by J. M. Gray.
by G. R. Sharp, April1978, 23 pp. IISBN: 1-58145-235-7). 12) A Review of the Structure and Propcrties of Welds in Columbium or Vanadium
235. lmprol'ed Repeatabílity in Ultrasonic Examination by A. S. Birks and W. E. Lawric, Containing High StrenJ(th Low Alloy Steels by E. Levine and D. C. Hill, February
and (21 Ultrasonic Testing System Standardization Requirements, February 1978, 1976. 35 pp. <ISBN: 1-58145-212-81
16 pp. tiSB!'\: 1-58145·234-9J. 212. Bibliography o{ Publu<hed Rep.>rts Resulting from the Work of WRC Project
234. Eflccts ofLack-o{Pcmetration and Lack-of-Fusion on the Fatigue Properties of5083 Committees. December 1975, 4 pp. tiSB!'\: 1-58145-211-X•.
Alumirwm Alloy Welds by J. D. Burk and F. V. LaY.-rence, Jr., January 1978, 14 pp. 211. Stress-Relie[Cracking rn St('(;i Weldment' by C. F. Meitzner, November 1975, 18 pp.
(ISB~: 1-58145-233-0). tiSBN: 1-58145-210-1
233. Report ofGasket Factor Tests by H. D. Raut and G. F. Leon. Deeember 1977,35 pp. 210. The Fabrication o{ Diss1milar Metal .]oínts Containing Reactive and Refractory
(ISBN: 1-58145-232-2). Metais by M. ~1. Schwartz, October 1975, 42 pp. 1ISBN: 1-58145-209-81.
232. Throuf(h Thickness Propcrties and ú:zmellar Tharing (a Biblicgraphy) by D. H. 209. Long-Range Plan for Pressure Vessel Reaearch-Fourth Edition by Pressure Vessel
Skinner and M. Toyama, November 1977, 20 pp. <ISB::\: 1-58145-231-4). Research Committee. September 1975, reprinted December 1976, 29 pp. <ISBN:
231. Factors J\ffectinf( 1\'C/d Metal Properties in Carbon and Low Alloy Pressure Vessel 1-58145-208-XI.
StPPls by K. F:. Don-chu, October 1977, 64 pp. <ISB~: 1-58145-230-6J. 208. Review o{ Data on M1tre Joints in Pipíng of Establish Mwcimum Angulanty for
230. An Experimental Study of Elasto-Plastic Response o{ Branch-Pl/x' The Connections Fabrication ofGirth Butt Welds by E. C. Rodabaugh, August 1975, 22 pp. IISBN:
Sub)Ccted to In tema/ Pressure, Extemal Couples and Combined Loadings by F. EUyin 1-58145-207-1).
and 21 Collapse Te..~/ of a Thin- Walled Cylindrical Pressure \k.~sel u·ith Radially 207. Joining o{ Metal-Matrix Fibcr-Reinfo~d Compo.~ite Materiais by G. L. :\1etzger,
Attached Nozzle by R. L. Maxwell and R. W. Holland, September 1977, 38 pp. IISBN July 1975, 22 pp. <ISBN: 1-58145-206-3
1-58145-229-2). 206. Effects of Porosity on thc Fatigue Properties o{ 5083 Aluminum Alloy Weldmcnts
229. Dynamic Fraclrll·t•-Resistance Testing and Methods for Structural Ana/ysis by E A. by F. V. Lawrence, Jr.. W. H Munse, and J. D. Burk, June 1975, 23 pp. <ISBN:
Lange, 12J Junction Stresses for a Conical Section Joining Cylinders of Different 1-58145-205-5 I.
Diametcr Subject to Intemal Pressure by W. J. Graff, August 1977, 18 pp. OSBN: 205. Fracture 1bughness and Related CharaC'Ieristics of the Cryogenic Nickel Steels, A.
1-58145-228-4 ). W. Pense nnd R. D. Stout, Mny 1975, 43 pp. (ISBN: 1-58145-204-7).
228. Statistica/ Analysi.~ ofDependence o{Weld Metal Properties on Composition by J. A. 204. Friction Welding by K K. Wang, April 1975, 21 pp. (ISBN: 1-58145-203-91.
.Marshall and J. Heu~chkel. July 1977, 34 pp. !ISBN: 1-58145-227-6> 203. Niobium. and Vanadwm ·Contain ing St~ls for Pressure \kssel Seroice by J. N. Cordea,
227. Test.~ o(Torispherical Pressure Vessel Heads Convex to Pressure by C. E. Washington, February 1975, 37 pp. tiSBN: 1-58145-202-0L
R. J. Clifton, nnd B. W. Costerus, and (2) Reference Stress Concepts for Creep Analysis 202. Current Welding Research Problems, Compiled by the Welding Research Council,
by H. Kraus, June 1977, 29 pp. IISB~: 1-58145-226-81. January 1975. 21 pp. <ISBN 1-58145-201-2).
226. A Ret•iett' of Minor Element Effects on the Welding Arr and Weld Penetration by S. 201. The Submerg<'d Are \leld in HSLA Line Pipe-A State-ofthe-Art Reuieur by P. A..
S. Glickstein and W. Yeni~cavich, May 1977, 18 pp. IISBN: 1-58145-225-XJ. Tichaurer, 12> Experiencl! in the D.:uelopment and Welding of Large-Diameter Pipes
225. Resi8tnncc Seam Welding by A W. Schueler, April 1977. 23 pp. IISBN: 1-58145-224-1). by 1\1. Civallero. C. Parrini, and G. Salmoni, (3) New Deuelopment in Weldability
and Welding Thchnique for Arctic-Grade Line Pipe by E. Miyoshi, Y. Ito, H. Iwanaga, 181. Effects of Porosity on the Tensile Properties of 5083 and 6061 Aluminum Alloy
and T. Yamura, (4) Thchnology of Wires and Electrodes for Welding High-Strength Weldments by F. V. Lawrence, Jr. and W. H. Munse, February 1973, 23 pp. (ISBN:
Pipe by J. Grosse-Wordemann, and (5) Preliminary Evaluation of Laser Welding of 1-5814.'5-180-6).
X-80 Arctic Pipeline Steel by E. M. Breinan and C. M. Banas, December 1974, 57 pp. 180. Elastic-Plastic Bending o{ a Constrained Circular Perforated Plate Under Uniform
OSBN: 1-58145-200-4). Pressure (Triangular Penetration Pattern) by J. S. Porowski and W. J. O'Donnell,
200. Analysis of Test Data on Branch Connections Exposed to Internai Pressure andl or January 1973, 17 pp. <ISBN: 1-58145-179-2).
Externai Couples by J. Schroeder, K R. Srinivasaiah, and P. Graham, November 179. Stress lndices and Flexibility Factors for Moment Lcadings on Elbows and Curued
1974, 26 pp. <ISBN: 1-58145-199-7). Pípe by W. G.DQdge and S. E. Moore, December 1972, 19 pp. (ISBN: 1-5814.'5-178-4).
199. Fatigue Behavior ofAluminum Alwy 5083 Butt Welds by W. W. Sanders, Jr. and S. M. 178. Joining Ceramics to Metais and Other Materiais by H. E. Pattee, November 1972,
Gannon, October 1974, 13 pp. <ISBN: 1-58145-198-9). reprinted May 1978, 43 pp. (ISBN: 1-58145-177-6).
198. Secondary Stress Indices for Integral Structural Attachments to Straight Pipe by W. 177. Comparison and Analysis of Residual Stress Measuring Thchniques, and The Effect
G. Dodge, and (2) Stress lndices at Lug Supports on Piping Systems by E. C. ofPost-Weld Heat Treatment on. Residual Stresses in Inconel 600, lnconel X -750 and
Rodabaugb, W. G. Dodge, and S. E. Moore, September 1974, 45 pp. (ISBN: 1-58145- Rene 41 Weldmen.ts by H. B. Peacock, C. D. Lundin, and J. E. Spruiell, October 1972,
197-0). 24 pp. fJSBN: 1-58145-176-8).
197. A Reuiew of Underclad Cracking in Pressure-Vessel Comporumts by I. G. Vinck.ier and 176. Long-Range Plan for Pressure- ~ssel Research-Third Edition by The Pressure Vessel
A. W. Pense, August 1974, 35 pp. (ISBN: 1-58145-196-2). Research Committee, September 1972, 23 pp. USBN: 1-5814.'5-175-X).
196. Electron Beam Welding by M. M. Schwartz, July 1974, 60 pp. (ISBN: 1-58145- 175. PVRC Recommendations on. 1bughness Requíremen.ts for Ferritic Materiais by The
195-4). PVRC Ad Hoc Task Group on Toughness Requirements, August 1972, 24 pp. <ISBN
195. A Review of Bounding Techniques m Shakedown and Ratcheting at Elevated 1-58145-174-1).
Thmperatures by F. A. Leckie, 12) A Reuiew of Creep Instability in High -Thmperature 174. Residual Stresses and Distortion in Welded Alumín.um Structures and Their
Piping and Pressure Vessels by J. C. Gerdeen and V. K Sazawal, (3) Upper Bounds Effects on Service Performance by K. Masubuchi, July 1972, 30 pp. <ISBN: 1-
for Accumu.lated Strains Due to Creep Ratcheting by W. J. O'Donnell and J. 58145-173-3).
Porowsk.i, and 14) Cyclic Creep-An lnterpretiue Literature Survey by E. Krempl, 173. Design o{ Tapered Members by G. C. Lee, M. L. MorreU, and R. L. Ketter, June
June 1974, reprinted March 1976, 123 pp. <ISBN: 1-5814.'5-194-6). 1972,32 pp. <ISBN: 1-58145-172-5).
194. Fatigue Behavior of Pressure- ~ssel Steels by J. M. Barsom, May 1974, reprinted 172. Sensitivity ofthe Delta Test to Steel Compositions and Variables by L. J. McGeady,
May 1978, 22 pp. OSBN: 1-58145-193-8). and (2) Experimental Stress Analysis and Fracture Behavior ofDelta Specimens by
193. Basic Considerations for Thbular Joint Design in Offshore Construction by P. W. J. M. Barsom, May 1972, 26 pp. (ISBN: 1-58145-171-7).
Marshall, April 1974, 18 pp. <ISBN: 1-58145-192-XJ. 171. Fatigue Behauior ofAluminum Alloy Weldments by W. W. Sanders, Jr., April 1972,
192. Allowable Stress for Web-Tapered Beams with Lateral Restraints by M. L. Morrell reprinted October 1976, 30 pp. (ISBN: 1-58145-170-9).
and G. C. Lee, February 1974, reprinted December 1976, 12 pp. (ISBN: 1-58145- 170. MIG Welding and Pulsed Power by A. Lesnewich, February 1972, 15 pp. (ISBN:
191-1}, 1-58145-169-5).
191. Submerged Are-Welding Procedures for Steels Meeting Standard Specifications by 169. Uniuersity Welding Research Directory, January 1972, 53 pp. (ISBN: 1-58145-
C. W. Ott and D. J. Snyder, January 1974. revised August 1987, 45 pp. (ISBN: 168-7).
1-58145-190-3) see Bulletin 326. 168. Lamellar Tearing by J. E. M. Jubb, December 1971, 14 pp. (ISBN: 1-58145-167-9).
190. Fluxes and Slags in We/ding by C. E. Jackson, December 1973, reprinted January 167. Laser Welding and Cutting by M. M. Schwartz, November 1971, reprinted March
1985, 25 pp. <ISBN: 1-58145-189-Xl. 1976, 34 pp. (ISBN: 1-58145-166-0).
189. Hardness as an lnde:r: of Weldability and Service Performance of Steel Weldments 166. Review of Service Experi~nce and Thst Data on Openings in Pressure Vessels with
by R. D. Stout, November 1973, 13 pp. (ISBN: 1-58145-188-1). Non-lntegral Reinforcing by E. C. Rodabaugh and (2) Derivation o{Code Formulas
188. Behavior and Design ofSteel Beam-to-Column Moment Connecticn.s by J. S. Huang, for Part B Flanges by E. O. Waters, October 1971, 37 pp. (ISBN: 1-58145-165-2).
w_ F. Chen, and L. S. Beedle and (2) Thst of a Fully-Welded Beam-to·Column 165. The 1bughness o{ 2% and 3'k Nicke/ Steels at Cryogenic Temperatures by N. J.
Connection by J. E. Regec, J. S. Huang, and W. F. Chen, October 1973, 35 pp. (ISBN: Huettich, A. W. Pense, and R. D. Stout, September 1971, 22 pp. <ISBN: 1-58145-
1-58145-187-3). 164-4).
187. High Temperature Brazing by H. E. Pattee, September 1973,47 pp. (ISBN: 1-58145- 164. Plastic Thsts of'Il.uo Branch-Pipe Connections by N. C. Lind, A. N. Sherbourne, F.
186-5). Ellyin, and J. Danora, (2) Bending of Pipe Bends with Elliptic Cross Sections by G.
186. Design Options for Selection of Fracture Control Procedures in the Modernization of E. Findlay and J. Spense, August 1971. 16 pp. <ISBN: 1-58145-163-6).
Codes, Rufes and Standards by W. S. Pellini, and (2> Analytical Design Procedures 163. Elastic-Plastic Deformations m Pressure ~ssel Heads by F. A. Simonen and D. T.
for Metais of Elastic-Plastic and Plastic Fracture Properties by W. S. Pellini, August Hunter, (2) Summary Report on Plastic Limit Analysis o{ Hemisph~rical-and
1973, 38 pp. <ISBN: 1-58145-185-7). Toriconical-Head Pressure ~ssels by J. C. Gerdeen and D. N. Hutula, July 1971,
185. lmproved Di.scontinuity Det~ction Using Computer-Aided Wtrasonic Pulse-Echo 32 pp. (ISBN: 1-58145-162-8).
Techniques by J. R. Frederick and J. A. Seydel, July 1973, 23 pp. (ISBN: 1-58145- 162. Unified Theory ofCumulative Damage in Metal Fatigue by J. Dudoc, B. Q. Thang,
184-9). A. Bazergui, andA. Biron, June 1971, 20 pp. CISBN: 1-58145-161-X).
184. Submerged-Arc-Weld Hardness and Cracking in Wet Sulfide Service by D. J. 161. The Fabrication and Welding o{ High-Strength Line-Pipe Steels by H. Thomasson,
Kotecki and D. G. Howden, June 1973, 22 pp. <ISBN: 1-58145-183-0). May 1971, 17 pp. (ISBN: 1-58145-160-1).
183. Criticai Literature Review ofEmbrittlement in 2CR-1 Mo Steel by L. G. Emmer, C. 160. High-Frequency Resistance Welding by D. C. Martin, April 1971, 24 pp. (ISBN:
D. Clauser, and J. R. Low, Jr., May 1973, 25 pp. <ISBN: 1-58145-182-2). 1-58145-159-8).
182. Brazed Honeycomb Structures by M. M. Schwartz, April 1973, 28 pp. (ISBN: 1-58145- 159. Welding of Maraging Steels by F. H. Lang and N. Kenyon, February 1971, 41 pp.
181-41. iiSBN: 1-58145-158-Xl.
442 Appendlx WRC Bulletlns 443
158. PVRC Interpretwe Report o{ Pressure Vessel Research, Section 3-Fabr~eation and 140. Fracture 1bughness Characterization Procedures and lnterpretatwns to Fracture·
Environmental C011.~iderations by A. P. Bunk, January 1971, 36 pp. (ISBN 1-58145- Safe Design for Structural Aluminum Alloys by R. W. Judy, Jr., R J. Goode, and C.
157-1). N. Freed, May 1969, 16 pp. <ISBN: 1-58145-139-3>.
157. Signifi.cance o{ Fracture Ertensíon Resístance (R Curve) Factors in FractuTT·Sa{e 139. Stress Concentrations in Th!o Normally lntersecting Cylindrical Shells Subject to
Design for Non{rangible Metais by W. S. Pellini and R. W. Judy. Jr.. December 1970, Internai Pressure by A. C. Erigen, A. K Naghdi, S. S. Mahmood, C. C. Thiel, and T.
20 pp. <ISB~: 1-58145-156-3 '· Ariman and <2 lA N?te on.the Correlation of Photoelastic and Steel Model Data for
156. Pilot Tests on the Static Strength o{ Unsymmetrical Plate Girders by J . R. Dimitri Nozzle Connectwns m Cylmdncal Shells by F. Sellars, April 1969, reprinted March
and A. Ostapenko, (2' Tests on a Transuersely Stiffened and on a wngitudinally 1976.39 pp. •ISBN: 1-58145-138-5).
Stiffened Unsymmetrical Plote Girder by W. Schueller andA. Ostapenko and (3) On 138. Intergron~lar Corr?sion of C~romium-Nickel Stainless Steels-Final Report by
the Fatigue Strength o{ Unsymmetrical Steel Plate Girders by P. Parsanejad andA. Subcom.mittee on Fteld Corroston Tests. February 1969, reprinted March 1976, 44
Ostapenko. November 1970. 59 pp. 1ISBN: 1-58145-155-5). pp. ·ISBN: 1-58145-137-7).
155. Fatigue and Static Tests o{ ThJo Welded Plate Girders by P. J. Patterson, J . A. 137. Current Welding Research Problems, January 1969, 30 pp. <ISBN: 1-58145-136-9).
Corrado, J. S. Huang. and B. T. Yen, Oct.ober 1970, 18 pp. (ISBN: 1-58145-154-7). 136. Creep-Rupture Properttes of Quenched and Tempered PressuTT 'kssel Steels-A
154. Electroslag. Electrogas, and Related Welding Processes by H. C. Campbell, Data Summary by J. J. deBarbadillo, C. D. Clauser, A. W. Pense, V. S. Robinson, C.
September 1970, 22 pp. !ISBN: 1-58145-153-9>. J. P. Steiner, and R. D. Stout, December 1968, 16 pp. <ISBN: 1-58145-135-0J.
153. Interpretive Report on Oblique Nozzle Connections in Pressure 'kssel Heads and 135. Cyclic Pressure Tests o{Full Size Pressure Vessels by A. G. Pickett and S. C. Grigory,
Shells Under Internai Pressure ÚJading by J. L. Mershon, (2) Elastic Stresses Near a November 1968, 48 pp. !ISBN: 1-58145-134-2).
Ske1ved Hole in a Flat Plate and Applications to Oblique Nozzle Attachments in Sht>lls 134. Procedures for Fracture Toughness Characterization and Interpretations to Failure·
by F. Ellyin, !3l Photoelastic Determination of the Stresses at Oblique Openings in Safe Design for Struclural TitaniUm Alloys by R. J. Goode. R. W. Judy, Jr. and R
Plates and Shells by M. M. Leven, and !4)A PhotoelasticAnalysts o{Oblique Cylinder W. Huber, October 1968, 17 pp. CISBN: 1-58145-133-4).
Intersections Subjected to Internai Pressure by R. Fidler, August 1970, 85 pp. (ISBN: 133. Proposed Reinforcement Design Procedure for Radial Nozzles in Spherical Shells with
1-58145-152-0). Internai Pressure (Phase Report No. 1) by R. L. Cloud and E. C. ROOabaugh and (2,
152. Influence o{Weld Dc(ects on the Mechanical Properties o{Aluminum Alloy Weldments Proposed Rein{orcement Design Procedure for Radial Nozzles in Cylindrical Shells
by A. W. Pense and R. D. St.out, July 1970, reprinted March 1976, 16 pp. <ISBN: with Internai Pressure (Phase Report No. 4) by E. C. Rodabaugb and R. L. Cloud,
1-58145-151-0). September 1968, reprinted March 1976, 33 pp. iiSBN: 1-58145-132-6).
151. Further Theon>tical 'Ih!atment ofPerforated Plates with Square Penetration Pattems 132. The Measurement o{ Delta Ferrite m Austenitic Stainless Steels by R. B. Gunia and
by W. J. O'Donnell, June 1970, 12 pp. tiSBN: 1-58145-150-4). G. A. Ratz, August 1968. reprinted March 1976, 22 pp. IISBN: 1-58145-131-8).
150. Recenl Studies o{Cracking During Postwelding Heat 'Ih!atment o{Nickel-Base Alloys 131. Are Plasmas for Joining, Cutting and Surfacing by R. L. O'Brien, July 1968,
!l)Evaluatmg the Resi.~tance o(Rene 41 toStrain-Age Cracking by R. W. Fawley and reprinted May 1978, 37 pp. !ISBN: 1-58145-130-X>.
M. Prager. t2l Variables Jnfluencing the Strain-Age Cracking and Mechanical 130. Advances in Fracture 1bughneS!I Charat"terization Procedures and in Quantitatit•e
Propertie.~ o{ Rene 41 and Related Alloys by J. S. Carlt.on and M. Prager. and !3) A lnterpretations to Fracture-Safe Design for Structural Steels by W. S. Pellini, Ma v
Mechanism for Cracking Durinl( Postwelding Heat 'Ih!atment of Nickel-Ba.-w Alloys 1968, reprinted March 1976, 46 pp. (ISB!'J: 1-58145-129-6.. ·
by M. Prager and G. Sines, May 1970,32 pp. tiSBN: 1-58145-149-0). 129. Elastic Stresses in Pre!lsure ~ssel Heads by H. Kraus, April 1968. 26 pp. ·ISBN:
149. Contrai of Distorlion and Shrinkage in Welding by K. Masubuchi, April 1970, 30 1-58145-128-8,.
pp. IISBN: 1-58145-148-21. 128. Welding o( Precipitation-Hardeníng Nickel-Base Alloys by M. Prager and C. S. Sh1ra,
148. [nteraction Curves for Sections Under Combined Bia.:cial Bend1ng and Axial Force by February 1968, 55 pp. <ISBN: 1-58145-127-XI.
S. Santathadapom and W. F. Chen, February 1970, 11 pp. !ISBN: 1-58145-147-41. 127. Girder Web Boundary Stn:sses and Fatigue by J. A. Mueller and B. T. Yen, Januarv
147. Transition-Temperatun> Data for Five Structural Steels by J. H. Gross, January 1968, reprinted May 1978, 22 pp. <ISBN: 1-58145-126-1 '· -
1970, reprinted March 1976, 14 pp. USBN: 1-58145-146-6). 126. Fmcture Development and Material Properties in PVRC-Penn State Pressure 'kssel
146. Structural Stability Design Prouisions-A Companson ofthe Provisions ofthe CRC by L. A. Cooley and E. A. Lange and (2) Fatigue-Crack Propagation and Fracture
Guide a11d the Specifi.cations o{AASHO, AISC, and AREA by B. T. Yen. J. S. Huang, Studies ofaPressure- Ve.~st>l Steel Temper Embrittled to Simulate lrradiation Damage
P. J . Patterson, and ,J. Brozzetti, November 1969, 34 pp. <ISBN: 1-58145-145-8~ by T. W. Crooker, L. A. Cooley, andE. A. Lange, November 1967, 11 pp. IJSBN:
145. Interpretive Report on Effect ofHydrogen in Pressure-Vessel Steels, Section l-Bas1c 1-58145-125-3).
and Research Aspecft< by C. G. lnterrante, Section II-Action o{ H_vdrogen on Steel 125. Analysis of In-Plane T. Y and K Welded Thbular Connectwns by L. A. Beale andA.
at High Temperature and High Pressun>s by G. A. Nelson and Section 1/I A. Toprac. October 1967, 30 pp. (ISBN: 1-58145-124-5).
Practical Aspects o( Hydrogen Damage at Atmospheric Temperature by C M. 124. A Study o{ Perforated Plates wtth Square Penetratton Patterns by W. J. O'Donnell,
Hudgins, Jr., October 1969, 52 pp <ISBN: 1-58145-144-X). September 1967, 13 pp. <ISBN 1-58145-123-7).
144. wng·Range Plan for Pressure-Vessel Research-Second Edition by Pressure Vessel 123. Fundamentais of Heat Flow in Welding by P. S. Myers, O. A. Uyehara, and G. L.
Research Committee, September 1969, 27 pp. USBN: 1-58145-143-1>. Bonnan, July 1967, reprinted March 1976, 46 pp. (ISBN: 1-58145-122-9).
143. Cast Heat-Resistant Alloys for High-Temperature Weldments by H. S. Avery,Augu~t 122. A Review o{ Some Microstructural Aspects of Fracture in Crystalline Materiais by
1969. 74 pp. <ISBN 1-58145-142-31. C.P. Sullivan, May 1967, reprinted March 1976, 57 pp.!ISBN: 1-58145-121-0).
142. Expenmcnts on Wide·Flange Beams Under Monumt Gradient by A. F. Lukey, R. J. 121. Residual Stresses in "T-1" Constructional Alloy Steel Plotes (2) Residual Stresscs
Smith, M . U. Hosain, and P F. Adams and (2) Variable Repeated Loading-A in Welded Bulld-Up "T-1" Shapcs, and (3) Residual Stresses in Rolled Heat-'[}eated
Literature Survey by D. G. Eyre and T. V. Galambos. July 1969, 26 pp. <ISBN: "T-1" Shapes by E. Odar, F. Nishino, and L. Tall, April1967, 36 pp. IISBN: 1-58145-
1-58145-141-5). 120-2>.
141. Integration of Metallurgical and Fracture Mechanics Concepts of Transition 120. The Properties and Microstructure o{ Spra_y.Quenched Thick-Sectwn Steels by S. S.
Temperature Factors Relating to Fracture-Safe Design for Structural Steels by W. Strunck, A. W. Pense, and R. D Stout. (21 Determination o( Fracture 1bughness o{
S . Pellini and F. J. Lo~s, June 1969, 38 pp. ISBN: 1-58145-140-7). Heavy-Sectwn PTTssure \'esse/ Ste.-/s Using a Fracture Mechanics Approach by E.
Landerman and S. E. Yanichko, and (3) Notch Properties o{ Some Low and Mediwn Electron Microscope Fractographic Study of Selected Portions ofthe Fracture Surface
Strength Con.structional Steels by G. Yoder, V. Weiss, and L. W. Liu, February 1967, by C. D. Beachem and E. P. Dahlberg, August 1964, 20 pp. (ISBN: 1-58145-
28 pp. (ISBN: 1-58145-119-9). 097-4).
119. 1nterpretiue Report on Pressure Véssel Heads by R. L. Cloud, and (2) 1nterpretations 97. D.C. Welding Power Sources for Gas-Shielded Metal-Are Welding by R. L. Hackman
o{Experimental Data on Pressure Vessel Heads Conuex to Pressure by R. J. Slember andA. F. Manz, July 1964, reprinted April1978, 22 pp. (ISBN: 1-58145-096-6).
and C. E. Washington, January 1967,49 pp. (ISBN: 1-58145-118-0). 96. Stresses Near a Cylindrical Outlet in a Spherical Véssel by E. O. Waters and (2) Effects
118. Fatigue Thsts o{ Large-Size Welded Plate Girders by B. T. Yen and J . A. Mueller, of Externai Loadings on Large Outlets in a Cylindrical Pressure Vessel by D. E.
November 1966, 25 pp. (ISBN: 1-58145-117-2). Hardenbergh and S. Y. Zamrik, May 1964, reprinted March 1976, 23 pp. (ISBN:
117. Static Tests on Longitudinally Stiffened Plate Girders by M. A. D'Apice, D. J. 1-58145- 095-8).
Fielding, and P. B. Cooper, October 1966, 35 pp. (ISBN: 1-58145-116-4). 95. PVRC 1nterpretive Report of Pressure Vessel Research, Section 1-Design Consider-
116. Long Range Plan for Pressure Vessel Research by Pressure Vessel Research ations by B. F. Langer, April 1964, 53 pp. (ISBN: 1-58145-094-X).
Committee, September 1966, 20 pp. (ISBN: 1-58415-115-6). 94. The Design o{Thick-Walled Closed-Ended Cylinders Based on Torsion. Data by B
115. Further Studies on the Lateral-1brsional Buckling of Steel Beam-Columns by T. V. Crossland, February 1964, 22 pp. (ISBN: 1-58145-093-1).
Galambos, P. F. Adams, and Y. Fukumoto, July 1966, 11 pp. (ISBN: 1-58145-114 -8). 93. Intergranular Corrosion of Chromium-Nickel Stainless Steel-Progress Report No.
114. Joining of Plastics by W. D. Harris, May 1966, 12 pp. (ISBN: 1-58145-113-X). 1 by Subcommittee on Field Corrosion Tests, January 1964, 25 pp. (ISBN: 1-58145-
113. Photoelastic Study ofthe Stresses Near Openings in Pressure Véssels by C. E. Taylor 092-3).
and N. C. Lind, (2) Photoelastic Determination ofthe Stresses inReinforced Openings 92. Brazed Honeycomb Structures by M. M. Schwartz and B. G. Bandelin, November
in Pressure Véssels by M. M. Leven, and (3) Preliminary Eualuation of PVRC 1963, 26 pp. (ISBN: 1-58145-091-5).
Phoroelastic Thst Data on Reinforced Openings in Pressure Véssels by J. L. Mershon. 91. Plastic Analysis and Tests of Haunched Corner Connections by J. W. Fi.sher, G. C.
April1966, reprinted March 1976, 70 pp. (ISBN: 1-58145-112-1). Lee, J. A. Yura, and G. C. Driscoll, Jr., October 1963, 33 pp. <ISBN: 1-58145-090-7).
112. 1mprouement ofthe Fatigue Life o{Spotwelds by J. A. Choquet, February 1966, 16 90. A Criticai Study ofthe Solution.s for the Asymmetric Bending of Spherical Shells, (2)
pp. <ISBN: 1-58145-111-3). Solutions for the Stresses at Nozzles in Pressure Véssels, and (3) Stress Concentration
111. Interpretiue Report on. Weld-Metal1bughness by K Masubuchi, R. E. Monroe, and Factors for the Stresses at Nozzle 1ntersections in Pressure Véssels by F. A. Leckie
D. C. Martin, January 1966, 38 pp. (ISBN: 1-58145-110-5). and R. K. Penny, September 1963, 26 pp. (ISBN: 1-58145-089-3).
110. Experiments on High Strength Steel Members by P. F. Adams, M. G. Lay, and T. V. 89. Experimental 1nuestigation ofStresses in Nozzles in Cylindrical Pressure Véssels by D.
Galambos, and (2) The Experimental Behauior ofRestrained Column.s by M. G. Lay E. Hardenbergh, S. Y. Zamrik, andA. J. Edmondson, July 1963, reprinted February
and T. V. Galambos, November 1965, 38 pp. (ISBN: 1-58145-109-1). 1977, 35 pp. (ISBN: 1-58145-088-5).
109. A Reuiew ofDiffusion Welding by J. M. Gerken and W. A. Owczarski, October 1965, 88. Fracture Analysis Di~gram Procedures for the Fracture-Safe Engineering Design of
28 pp. (ISBN: 1-58145-108-3). Shell Structures by W. S. Pellini and P. P. Puzak, May 1963, 28 pp. (ISBN: 1-58145-
108. Experimental Determination of Stress Distribution.s in Thin-Walled Cylindrical and 087-7).
Spherical Pressure Vessels with Circular Nozzles by W. F. Riley, (2) A Reuiew and 87. Criticai Factors in the lnterpretation. of Radiation E{fects on the Mechanical
Eualuation ofComputer Programs for the Analysis of Stresses in Pressure Véssels by Properties of Structural Metals by R. G. Berggren, (2) Commercial 1mplications of
H. Kraus, September 1965, 28 pp. (ISBN: 1-58145-107-5). Radiation Effects on Reactor Pressure Véssel Design, Fabrication and Operation by
107. Local Stresses in Spherical and Cylindrical Shells Dueto Externai Loadings by K. L. R. Weissert and D. K. Davies, (3) Design Criteria for 1rradiated Reactor Vessels
R. Wichman, A. G. Hopper, and J . L. Mershon, August 1965, March 1979 revision, by D. W. McLaughlin, and (4) Surueillance of Criticai Reactor Components to Assess
69 pp. (ISBN: 1-58145-106-7). Radiation Damage by L. E. Steele and J. R. Hawthorne, April 1963, 21 pp. (ISBN:
106. Stress Analysis o{ a Circular Plate Containing a Rectangular Array of Holes and (2) 1-58145-086-9).
Stress Distribution Around Periodically Spaced Holes in a Spherical Membran.e Shell 86. Welding of Steel at Low Ambient Temperatures by K. Winterton, W. P. Campbell,
Un.der Uniform 1ntemal Pressure by J. B. Mahoney and V. Salerno, July 1965, and M. J. Nolan, March 1963, 27 pp. (ISBN: 1-58145-085-0).
reprinted March 1976, 28 pp. (ISBN: 1-58145-105-9). 85. Welding of Reactive and Refractory Metais by E. G. Thompson, February 1963, 30
105. Are Spot Weldin.g by T. W. Shearer, May 1965, 27 pp. <ISBN: 1-58145-104-0). pp. USBN: 1-58145-084-2).
104. Bonding ofMetais with Explosives by A. H. Holtzman and G. R. Cowan, April 1965, 84. An Experimental 1nuestigation of the Stresses Produced in Spherical Véssels by
21 pp. (ISBN: 1-58145-103-2). Externai Loads Transferred by a Nozzle by J. W. Dally, January 1963, 29 pp. (ISBN.
103. Welding ofAge-Hardenable Stainless Steels by F. G. Harkins, February 1965, 25 pp. 1-58145-083-4).
(ISBN: 1-58145-102-4). 83. Are Welding ofMagnesium and Magnesium Alloys by C. R. Sibley, November 1962,
102. State of Stress in a Circular Cylindrical Shell with a Circular Hole by A. C. Eringen, 16 pp. <ISBN: 1-58145-082-6).
A. K. Naghdi, and C. C. Thiel, January 1965, reprinted March 1976, 21 pp. (ISBN: 82. Are Welding ofThick Cross Sections by R. E. Lorentz, Jr. and P. O. Leach, October
1-58145-101-6). 1962 17 pp. (ISBN: 1-58145-081-8).
101. PVRC lnterpretive Report of Pressure Vessel Research, Section 2-Materials 81. A Survey of Literature on the Stability of Frames by L.-W. Lu and (2) Tests on the
Considerations by J. H. Gross, November 1964, 31 pp. (ISBN: 1-58145-100-8). Stability of Welded Steel Frames by Y.-C. Yen, L.-W. Lu, and G. C. Driscoll, Jr.,
100. Electron Beam Welding by K. J. Miller and T. Takenaka, October 1964, 23 pp. (ISBN: Sept~mber 1962, 24 pp. (ISBN: 1-58145-080-XJ.
1-58145-099-0). 80. Analysis of a Perforated Circular Plate Containing a Rectangular Array o f Holes by
99. Experiments on Braced Wul.e-Flange Beams by G. C. Lee, A. T. Ferrara, and T. V. J. B. Mahoney, V. L. Salerno, and M. A. Goldberg and (2) Stresses and Deflections
Galambos and (2) The Experimental Bases for Plastic Design by M. G. Lay, September in Laterally Loaded Perforated Plates by D. Bynum, Jr. and M. M. Lemcoe. August
1964, 32 pp. (ISBN: 1-58145-098-2). 1962, 25 pp. !ISBN: 1-58145-079-6).
98. Failure Analysis ofPVRC Vessel No. 5, Part 1-A Study ofMateriais Properties and 79. Welding mtra High Strength Steel Sheet by H. Schwartzbart and J. F. Rudy, July
Fracture Deuelopment by E. A. Lange, A. G. Pickett, and R D. Wylie, and Part 11-An 1962, 17 pp. !ISBN: 1-58145-078-8).
78. Nonwgraphs for the Solution of Beam-Column Problerns by M. Ojalvo and Y. 56. lnterpretiue Report on Welding Titanium and 1ítanium Allvys by G. E. Faulkner
Fukumoto and (2) mtimate Strength Tables for Beam-Columns by T. V. Galambos and C. B. Voldrich, December 1959, 20 pp. (ISBN: 1-58145-055-9).
and J. Prasad, June 1962, 26 pp. (ISBN: 1-58145-077-X). 55. Fourth Technical Progress Report ofthe Ship Structure Committee, November 1959,
77. PVRC Research on Reinforcement ofOpenings in Pressure Vessels by J. L. Mershon, 14 pp. (ISBN: 1-58145-054-0).
May 1962, reprinted March 1976, 54 pp. (ISBN: 1-58145-076-1). 54. Bending of2:1 and 3:1 Open-CrownEllipsoidal Shells by G. D. Galletly, October 1959,
76. Gas and Oil Pipeline Welding Practices by A. G. Barkow, April1962, 32 pp. (ISBN: 9 pp. (ISBN: 1-58145-053-2).
1-58145-075-3). 53. Plastic Design of Pinned-Base •Lean To" Frames by R. L. Ketter and B.-T. Yen,
75. Aluminum and Aluminum Alloys for Pressure Vessels by M. Holt, J. G. Kaufman, September 1959, 20 pp. (ISBN: 1-58145-052-4).
and E. T. Wanderer, February 1962, 31 pp. (ISBN: 1-58145-074-5). 52. A Review, Comparison and Modification of Present Deflection Theory for Flat
7 4. A Critica/ Evaluation of the Strength of Thick-Walled Cylindrical Pressure Vessels Perforated Plates by V. L. Salemo and J. B. Mahoney and (2 ) Correlation of
by J. Marin and T.-L. Weng, January 1962, 15 pp. (ISBN: 1-58145-076-1). Experimental Data with Theory for Perforated Plates with a Triangular Hole Array
73. Copper and Copper Alloys for Pressure Vessels by V. P. Weaver and J. Imperati , by L. Deagle, July 1959, 27 pp. (ISBN: 1-58145-051-6).
November 1961, 21 pp. (ISBN: 1-58145-072-9). 51. Theoretical Stresses Near a Circular Opening in a Flat Plate Reinforced with a
72. Thermal Fatigue-A Criticai Review by T. C. Yen, October 1961, 12 pp. (ISBN: Cylindrical Outlet by E. O. Waters, (2) Stresses in Contoured Openings of Pressure
1-58145-071-0). Vessels by D. E. Hardenbergh, (3) A Three-Dimensional Photoelastic Study of
71. Research on Thbular Connections in Structural Work by J. G. Bouwkamp and (2)An Stresses Around Reinforced Outlets in Pressure Vessels by C. E. Taylor, N. C. Lind,
Investigation of Welded Steel Pipe Connections by A. A. Toprac, August 1961, 33 pp. and J. W. Schweiker, and (4) Unreinforced Openings in aPressure Vessel by F. S. G.
(ISBN: 1-58145-070-2). Williams andE. P. Auler, June 1959, 46 pp. (ISBN: 1-58145-050-8).
70. lnert-Gas-Shielded Are Welding of Ferrous Metal by G. R. Rothschild, and A. 50. Stresses in Spherical Vessels from Local Loads Transferred by a Pipe and (2}
Lesnewich, July 1961, 21 pp. (ISBN: 1-58145-069-9). Additional Data on Stresses in Cylindrical Shells under Local Loading by P. P.
69. The Effects ofInternai Pressure on Thin-Shell Pressure Vessel Heads by E. O. Jones, Bijlaard, May 1959, reprinted March 1976, 50 pp. (ISBN: 1-58145-049-4).
Jr. with Discussion by L. P. Zick, (2) Wrinkling of a Large Thin Code Head Under 49. Stresses in a Spherical Vessel from Radial Loads Acting on a Pipe, (2) Stresses in a
Internai Pressure by A. Fino and R. W. Schneider, and (3) Biaxial Stress Criteria for Spherical Vessel from Externai Moments Acting on a Pipe, and (3) lnfluence of a
Large Low-Pressure Tanks by J. J. Dvorak and R. V. McGrath, June 1961, reprinted Reinforcing Pad on the Stresses in a Spherical Vessel Under Local Loading by P. P.
March 1976, 24 pp. <ISBN: 1-58145-068-0). Bijlaard, April1959, 73 pp. (ISBN: 1-58145-048-6).
68. A Coated Electrode for Fusion Welding AIS! 4340 Steel for mtra High Strength 48. Plastic Design of Pinned-Base Gable Frames by R. L. Ketter, March 1959, 20 pp.
Applications by E. F. Deesing and (2) Method for Determining the 1btal Water in (ISBN: 1-58145-047-8).
Welding Electrode Coatings by K P. Johannes, April 1961, 15 pp. (ISBN: 1-58145- 47. An Experimentallnuestigation of Open-Web Bearns by A. A. Toprac and B. R. Cooke,
067-2). February 1959, 16 pp. (ISBN: 1-58145-046-X).
67. Strength ofThick-Walled Cylindrical Vessels Under Intemal Pressure for Three Steels 46. Observations of Strains Near Reinforced and Nonreinforced Cone Cylinder lnter-
by J. Marin and T.-L. Weng, March 1961, 13 pp. OSBN: 1-58145-066-4). sections by C. Kientzler and S. F. Borg and (2) Discussion "Design Formulas for a
66. lnterpretive Report on Operating Conditions for Nuclear Pressure Vessels by K F. Thin Cylinder with Cone Shaped Ends" by C. O. Rhys, January 1959, 10 pp. (ISBN:
Smith, January 1961, 12 pp. (ISBN: 1-58145-065-6). 1-58145-045-1).
65. Feasibility Studies ofStresses in Ligaments by M. M. Lemcoe, November 1960, 27 pp. 45. Ten Years of Progress in Pressure Vessel Research by F. L. Plummer, R. D. Stout, E.
(ISBN: 1-58145-064-8). Wenk, Jr., and I. E. Boberg, December 1958, 16 pp. (ISBN: 1-58145-044-3).
64. Web Buckling Thsts on Welded Plate Girders by K. Basler, B. T. Yen, J. A. Mueller, 44. The lnfluence of Residual Stress on the Strength of Structural Members by R. L.
and B. Thürlimann, September 1960, 64 pp. (ISBN: 1-58145-063-X). Ketter, November 1958, 11 pp. (ISBN: 1-58145-043-5).
63. Welded Interior Beam-to·Column Connections by J. D. Graham, A. N. Sherboume, 43. Welding of 347 Stainless Steel Piping and Thbing by G. E. Linnert, October 1958,
and R. N. Khabbaz under the direction of C. D. Jensen, (2) Transfer of Stresses in 103 pp. <ISBN: 1-58145-042-7).
Welded Couer Plates by A. M. Ozell andA L. Conyers, and (3) A Survey of Literature 42. Weld Flaw Eualuation by S. T. Carpenter and R. F. Linsenmeyer, September 1958,
reprinted February 1977, 37 pp. (ISBN: 1-58145-041-9).
on the Lateral lnstabilityofBeams by G. C. Lee, August 1960,59 pp. <ISBN: 1-58145-
41. Design of Thick-Walled Pressure Vessels Based Upon the Plastic Range by J . Marin
062-1).
and F. P. J. Rimrott, July 1958, 18 pp. (ISBN: 1-58145-040-0).
62. Comparison and Analysis of Notch 1bughness Tests for Steels in Welded Structures
40. Design of a Thin-Walled Cylindrical Pressure Vessel Based Upon the Plastic Range
by H. H . Johnson and R. D. Stout, July 1960, 28 pp. IISBN: 1-58145-061-3).
and Considering Anisotrophy by J. Marin and M. G. Sharma, May 1958, 13 pp.(ISBN:
61. Interpretive Report on Welding of Nickel-Clad and Stainless-Clad Steel Plate by W 1-58145-039-7}.
H . Funk, June 1960, 20 pp. (ISBN: 1-58145-060-5). 39. Plastic Analysis and Design of Square Rigid Frame Knees by J. W. Fisher, G. C.
60. An Experimental Investigation of Stresses in the Neighborhood ofAttachments to a Driscoll, Jr., and L. S. Beedle, April 1958, 8 pp. (ISBN: 1-58145-038-9).
Cylindrical Shell by E. T. Cranch and (2) Interpretiue Commentary on the Application 38. Proof'Thsting Pressure Vessels Designed for Interna[ Pressure by R. W. Schneider,
ofTheory to Experimental Results by P. P. Bijlaard andE. T. Cranch, May 1960, 44 July 1957, 8 pp. (ISBN: 1-58145-037-0).
pp. OSBN: 1-58145-059-1). 37. The Value of the Notch Tensile Test by J. F. Baker and C. F. Tipper, June 1957, 31
59. Factors WhichAffect Low-Alloy Weld Metal Notch-1bughness by S. S. Sagan and H. pp. (ISBN: 1-58145-036-2).
C. Campbell, April 1960 16 pp. (ISBN: 1-58145-058-3). 36. Ultra-High-Strength Weld Metal with Low-Hydrogen Electrodes by D. C. Smith,
58. Strength of Aluminum Alloy 6061-T4 Thick-Walled Cylindrical Véssels Subjected to May 1957, 14 pp. (ISBN: 1-58145-035-4).
Internai Pressures by J. Marin and T.-L. Weng, March 1960, 12 pp. (ISBN: 1-58145- 35. The Tensile Properties ofSelected Steels as a Function ofTemperature by E. P. Klier,
057-5). April1957, 9 pp. (ISBN: 1-58145-034-6).
57. Some Observations on the Brittle Fracture Problem by G. M. Boyd, January 1960, 34. Computation of the Stresses from Local Loads in Spherical Pressure Véssels or
9 pp. <ISBN: 1-58145-056-7). Pressure Vessel Heads by P. P. Bijlaard, March 1957, 8 pp. (ISBN: 1-58145-033-8).
33. Further Studus ofthe Hot-Ductility o{High-Thmperature Al{qys by E. F. Nippes, W. 6. Stress Corrosion Crackmg of Stainle.<!l Steels by H. Nathorst, Part I. Practical
F. Savage, and G. Grotke, February 1957, 32 pp. (ISBN: 1-58145-032-Xl. Experiences, Part li An lnvestigation of the Suitabil1ty of the U-bend Specimen,
32. Graphitization o{ Steel in Petroleum Refining Equipment and The Effect o{ October 1950. 18 pp. ISBN: 1-58145-005-2).
Graphitization of Suel on Stress-Rupture Propertus by J. G. Wilson, January 5. Hotc Plastic Deformation lnfluences Desígn and Forming of Metal Parts by J. R.
1957, 44 pp. <ISBN: 1-58145-031-1). Low. Jr. and (2) Brittle Fracture in Mild Steel by J. S. Hoggart, May 1950, 20 pp.
31. Stainless Steel for Pressure Vessels by A. Grodner, November 1956, 20 pp. <ISBN: ISBN: 1-58145-004-4).
1-58145-030-3). 4. Testing Pressurr: Ve.•.•els by F. G. Tatnall and (21 Effect o{Welding on Pressure Vessel
30. Oxidation-Resistant Brazing Al{qys by G. H. Sistare. Jr. and A. S. McOonald, Steels by A. F. Scotchbrook. L. Eriv, R. O. Stout, and B. G. Johnston, Februarv
September 1956, 13 pp. <ISBN: 1-58145-029-X). 1950, 12 pp. 'ISBN: 1-58145-003-6!. -
29. Deuelopment of Brazing Alloys for Joining Heat Resistant Alloys by F. M. Miller. H. 3. Burn·Off Charactenstics of Ste1.'l Welding Electrodes by O. C. Martin, P. J. Rieppel.
S. Gonser, and R. L. Peaslee, July 1956, 11 pp. <ISBN: 1-58145-028-1). and C. B. Voldrich, ~1ay 1949. 8 pp. <ISBN: 1-58145-002-8).
28. Aluminum and Aluminum Alloys for Pressure ~ssels by M. Holt, June 1956, 24 pp. 2. The .'Vature ofthe Are by J. O. Cobine, April 1949. 6 pp. 'ISBN: 1-58145-001-X
(ISBN: 1-58145-027-3). 1. Steel Compositions and Specifications by C. M. Parker, March 1949. 11 pp. ISBN'
27. Mechanical Properties and Weldability ofSix High-Strength Sreels by R. O. Stout and 1-58145-000-1}.
J. H. Gross, and (2) Economic but Safe Pressure-Vessel Construction by J. J. Murphy,
C. R. Soderberg, Jr., andO. B. Rossheim, May 1956,39 pp. (ISBN: 1-58145-026-5).
26. /nuestigation ofthe Influence of Deoxidation and Chemical Composition on Notched-
Bar Properties of Ship Plate Sreels by F. W. Boulger, R. H. Frazier, and C. H. Lorig,
April1956, 18 pp. (ISBN: 1-58145-025-7).
25. Bibliography on the Welding o{Stainless Steels, 1926-1955 with Author lndex by K.
Janis, March 1956, 27 pp. (ISBN: 1-58145-024-9).
24. Nickel and High-Nickel Alloys for Pressure Vessels by R. M. Wilson, Jr. and W F.
Burchfield, January 1956, 27 pp. (ISBN: 1-58145-023-0).
23. Cracking ofSimple Structural Geometries by S. T. Carpenter and R. F. Linsenmeyer,
July 1955, 24 pp. (ISBN: 1-58145-022-2).
22. Welded The Connections by A. G. Barkow and R. A. Huseby, May 1955, 24 pp.
(ISBN: 1-58145-021-4).
21. Coating Moisture lnuestigations of Ausrenitic Electrodes of the Modified 18-8 Type
by K. P. Johannes, O. C. Smith, and W. G. Rinehart, February 1955, 18 pp. (ISBN·
1-58145-020-6).
20 High-Strength, Low-Alloy Steels; Analytical Chemistry Fosters Progres~; in Steel-
making; Steel Quality by C. M. Parker, January 1955, 16 pp. <ISBN: 1-58145-019-2 .
19. Review o{ Welded Ship Failures by H. G. Acker, November 1954, 15 pp. <ISBN:
1-58145-018-4).
18. Thsting of Stainless-Steel Weldments by H. Thielsch, October 1954, 26 pp. <ISBN
1-58145-017-6).
17. A Criticai Survey o{ Brittle Failure in Carbon Plate Steel Structures Other Than
Ships by M. E. Shank, January 1954, 48 pp. iiSBN: 1-58145-016-8).
16. Third Thchnical Progress Report of the Ship Structure Committee, and Research
Under the Ship Structure Committee by Captain E. A. Wright, F. Jonassen, and H .
G. Acker, November 1953, 64 pp. (ISBN: 1-58145-015-X).
15. Slag-Metal lnteraction in Are Welding, N. Christensen and J. Chipman, January
1953, 14 pp. <ISBN: 1-58145-014-1).
14. Welding Processes and Procedures Employed in Joining Stainles!l Steels, H.
Thielsch, September 1952, 48 pp. (ISBN: 1-58145-013-3).
13. A New High· Y"Leld Strength Alloy Steel for Welded Structures by L. C. Bibber, J. M.
Hodge, R. C. Altman, and W. O. Ooty, July 1952, 17 pp. (ISBN: 1-58145-012-5).
12. An Analytical Study of Aluminum Welding by C. O. Smith, E. R. Funk, and H.
Udin, June 1952, 8 pp. (ISBN: 1-58145-011-7).
11. /nvestigations of Effect o{ Fabrication Operations Upon Pressure ~ssel Steels by S.
S. Tor and R. O. Stout, May 1952, 12 pp. (ISBN: 1-58145-010-9).
10. Thermal Fatigue and Thermal Shock by H. Thielsch, April1952, 24 pp. (ISBN:
1-58145-009-5).
9. Copper in Stainless Steels by H. Thielsch,August 1951,31 pp. (ISBN: 1-58145-008-7).
8. Thsts o(Columns Under Combmed Thrust and Moment by L. S. Beedle, J. A. Ready,
and B. G. Johnston, Oecember 1950, 23 pp. (ISBN: 1-58145-007-9).
7. lnstrumentation for the Evaluation ofthe Stabüity ofthe Welding Are by L. P. Wmsor,
L. M Schetky, and R. A. Wyant, November 1950.9 pp. ·ISBN: 1-58145-006-0).
lndex
A-scan, 244 B-scan, 244

AASHTO, 100, 108 Bending:
Abnormal loads, 71 accidental, 292
Acid wash, 199 cold, 73
Acoustic emission, 248 failure, 373
Allowable stress, 84 stress, 74-75, 93
Alloy steel, 28, 35 Biocides, 177
Anchor motion, 370 BLEVE, 227, 236
Anode, 157-160 Blister, 188, 190-192, 300-302
sacrificial, 196 Blowout, 127
API: Boat sample, 51
API 510, 17, 401, 404 Body centered cubic, 38-39
API 530, 336, 343 Bolted joint, 357
API 570, 17, 402, 412 Borescope, 214
API579, 16-18,103,254,275,314, Branch, 127
320,333,347 Brinell hardness, 64
API 580,217 Bubble collapse, 357
API 581,217, 232 Buckle, 258, 294, 298
API 620, 17,69 Burn-through, 127, 141
API 650, 17, 69, 133, 260 Burst, 9, 80-82
API 653, 17, 262, 416
API 2201, 404 C-scan, 244
Are strike, 141 Carbon, equivalent, 124
ASME III NH, 337 Carburization, 192
ASME XI, 402, 404 Cast iron, 26
ASME B31: Casting quality factor, 396
B31.1, 266, 277, 393 Cathodic protection, 199
B31.3, 68, 91, 269, 346 Cavitation, 179
B31.4, 96, 393, 412 Charpy toughness, 35, 59, 275
B31.8, 97, 393 Clamp, 397
B31.8S, 249 Coal-tar enamel, 196-197, 202
B3l.G, 238, 270-276 Coating, 174, 196-208
pressure piping code, 68-69 multilayer, 203
ASME Code sections, 17,67- 68 quality control, 205
Asphalt coating, 196 Collapse, 113, 372
ASTM coating, 206 Cold spring, 73
Auger electron spectroscope, 192 Corrective maintenance, 11
Austenitic stainless steel, 29 Corrosion, 153
AWS, 100, 108 carbon diox:ide, 178
AWWA,69 concentration cell, 161
451
452 lndex lndex 453
Corrosion (Cont. ): Dislocation, 35 Forging, 79 J et fire, 227

crevice, 170 Drop-weight test, 35 Forming, 120 Jurisdictional authority, 7
fatigue, 186 Ductility, 25, 46, 52 Fracture:
future corrosion allowance, 69 Duplex stainless steel, 31, 314 brittle, 55-57 Knoop hardness, 64
galvanic, 160--161, 169 ductile, 53
general, 167, 254 Eddy current, 145,246 toughness, 60--62 Lamination, 118
high-temperature, 192-194, 331 Elasticity, 74 Fracture mechanics, 100, 106 Lap, 120
liquid-line, 175 Electrical resistance, 195 Fracture toughness, 311, 314-315 Larson-Miller, 45, 340--341
microbial induced, 175-178 Electrochemical: Fragmentation, 365 Leak:
parameters, 163-165 cell, 157-160 Freeze plug, 403 box,413
pitting, 169 impedance, 195 Fusion bonded epoxy, 198 or break, 254, 382
rate, 164-166, 168, 194 Elliptical, 87 detection, 249
sour, 188 Embrittlement, 193 Galvanic, 154, 160--161 pinhole, 257
under insulation, 173 Energy, 388 General metalloss, 253 tightness, 12
Coupon, 195 Energy dissipative X-ray, 381 Geometry factor, 313 Lemont refinery, 189
Crack, 106-108 Epoxy, 197, 200-202 Glass fiber, 196 Length of thickness averaging,
blister, 190--192 EPRI, 332 GML, 253 256, 267
in dent, 297 Erosion, 179, 181-184 Gouge, 119,258,294,296 Ligament stress, 315
environmental, 184 Expansion, thermal, 73, 94 Graphitic corrosion, 192 Limit state function, 228
FAD, 324 Explosion, 83-84, 305, 360, 363 Graphitization, 192, 193, 331 Linear polarization resistance, 195
growth, 107 Externai explosion, 368 Guided wave, 246 Liquid penetrant testing, 241
opening area, 325 Extruding, 120 Load:
prevention, 129 Half-life, 395 expansion, 93
RBI, 230 Face-centered cubic, 37 Hammer: occasional, 93
stability, 310 Failure assessment diagram, bubble collapse, 356 operating, 337
weld, 129, 138-139 317-319,389 liquid hammer, 71 overload, 99
Creep, 329-350 Fatigue, 83, 97-110 waterhammer, 355-357 sustained,93
definition, 329 corrosion, 186 Hardness,62, 148,409 Local metalloss, 253
life, 340, 247 curve, 103 Hastelloy, 179 Lot, 45
metallographic, 349 fracture mechanics, 326 Heat, 45 LTA,253,279
stages, 329 strength reduction factor, 100 Heat treatment, 39-45
Criticai thickness profile, 255, 268, usage factor, 103 Hemispherical, 86 Magnetic flux leakage, 246
269 Ferritic steel, 30 HIC, 190--192, 409 Magnetic particle testing, 240
Cycles, 101-102 Ferrous, 26 High-temperature Magnetite, 333
Fiberglass, 422 corrosion, 154 Maintenance, 378
Decarburization, 192 Fillet welded patch, 414 Role drilling, 145 Management of change, 6
Defect, 118 Fitness-for-service, 1, 6 Hydrogen beak-out, 409 MAOP, 312
Deflagration, 360, 364 Fitting, 79-80, 93 Hydrogen blister, 190 Margin, 11, 323
Deformation, 83 Flange, 79 Hydrostatic, 134 Markl method, 99, 104, 110
Dent, 294-297 Flashfire, 227,236 Hydrotest, 384 Martensite, 37-38
Deposit, 166 Flat head, 89 Hydroxide, 158 Martensitic steel, 30
Design: Flaw: Materiais properties council, 147, 332
basic, 67 excavation, 407 Impact, 95 Mechanical clamp, 417
detailed, 67 fabrication, 117 Inclusion, 8, 120, 140 Mechanical properties, 25
minimum wall, 69 handling, 137 Inertia, 369 Microstructure, 35
principie, 82 mill, 137 Intergranular, 184 Misalignment, 141
Detonation, 361-363 Flow stress, 228, 366 Inserted liner, 418 Monel, 179, 188
Dillon, C.P., 153-154 Flush patch repair, 406 Inspection, risk-based, 213 Monkman-Grantconstant,348
Discontinuity, 98, 100 Fontana, M.G., 153 Ion, 157-158, 162 MTBF, 215
454 lndex lndex 455
NACE, 165, 198 Pressure Vessel Research Council, 147 Steam: TOFD, 245
Natural phenomena hazards, 369 Preventive maintenance, 11 line, 78 Tolerance, 77
NBIC, 17,402 RBI,223-226 Torispherical, 89
Nil ductility transition, 58-59 Qualitative RBI, 219 stored energy, 384 Tbughness,25,34,58, 188
Nondestructive examination, Quantitative RBI, 220 Steel, 27 Transient flow, 71
213 Stellite, 179 Transmission electron microscope, 381
Nonferrous, 31-32 Radiographic testing, 242, 345 Stored energy, 388 Tuberculation, 166--167
Nonmetallic, 33 Reagent, 159 Strain,82,364,367-368
Normal operating loads, 71 Reference stress, 315, 322 Strength,25,34,47-49 UVCE, 236
Regulatory, 14 Stress: Ultrasonic testing, 243, 345
0-liner, 418 Rehabilitation, 207 allowable, 49-50 Undercut, 141
Operability, 377 Reliability, 232 alternating, 101
Optical microscope, 381 Repair, 393-423 amplitude, 101, 104 Vacuum, 112
Overheating, 335 Replacement, 405 bending, 74-75,93 Vapor cloud explosion, 227
Overlap, 140 Replication, 381 classification, 110-112 Vibration, 105
Overload, 351-374 Residual stress, 73, 99, 143--149, 315, correction, 102 Vickers, 64
Overpressure, 83 320,322 cyclic, 98 Vinyl, 197
allowance, 351 Ripple, 300 engineering, 4 7 Visual examination, 239
considerations, 354 Riser, 292-294 expansion, 94 Vortex, 105
failure, 257 Risk, 216 gasket seating, 359
protection, 67 Risk-based inspection, 216-238 hoop, 76,255,289,344 X-ray diffraction, 145
Owner, 7 Rockwell hardness, 63 intensification, 94, 105
Rolling, 120 intensity, 101, 311-312 Yield, 96, 4 7, 52
Partial safety factor, 321 Root cause failure analysis, 378 longitudinal, 76 Young's modulus, 275, 325, 338
Passivate, 160 Roughness, 198 primary, 102
Permanent repair, 395 RSTRENG, 238, 276-277 range, 101-102 Wall thickness:
Phase diagram, 36 Rupture disk, 353 secondary, 102 creep, 348
Phenolic, 197 Rust, 160 self-equilibrate, 149 minimum, 77, 86
Pipe splitting, 419 true, 47-49 ordered, 77-79
Pipeline: Sag,83 Stress corrosion cracking, required, 77
assessment, 20 Scab, 119 184-185 Wall thinning, 156
crack, 312-319 Scaling, 334 Structural discontinuity, 258 Weld:
direct assessment, 249 Scanning electron microscope, 381 Surface finish, 98 autogenous, 121
integrity program, 238 Scleroscope, 64 branch, 99
pig,248-249, 294,360--361 Section modulus, 75, 93 Tank: controlled deposition, 398-402
Pitting, 169, 189, 253, 281-286 Sensitized, 186 collapse, 113-114 decay, 140
Plastic analysis, 372 Slag, 142 distortion, 291 distortion, 302
POD, 214 Sleeve: frangible roof, 386 fatigue, 109
Polyester, 197 full encirclement, 411 repair, 416 fillet, 100
Pool fire, 227, 236 heat-shrinkable, 203 top, 386 flaw, 137-143
Predictive maintenance, 11 SMYS,96-97 wall thinning, 259-265 heat treatment, 398
Pressure: SNT-TC, 214 Temper, 78 joint efficiency, 86, 91, 396
boundary, 11, 377 SOHIC, 190-192 Temper bead, 129 misalignment,302
design, 83 Soil settlement, 73 Temper embrittlement, 192 offset, 121
externai, 112 Sour corrosion, 188 Temporary repair, 395 overlay, 409
rating, 396 Spherical head, 86 Test: peaking,302
relief device, 83 SSPC, 198-199 leak, 130- 133 plug, 416
stress, 76 Stability, 377 pressure, 130-133,384 procedure, 125
test, 133-136 Stabilized steel, 187 proof, 396 qualification, 126, 128
456 lndex
Weld (Cont.): White metal, 198

quality. 125 Workmanship, 7
size, 143 Wrap tape, 199
spiral, 121 Wrap,no~etalüc,419
Wrinkle, 298 ABOUT THE AUTHOR
temper bead. 400
Welding techniques, 121-124 George Antaki is a Fellow of the American Society of
in senrice, 126-130, 398 Zero resistance ammetry, 195 ~1echanical Engineers. He is a member of the ASME
B31 Mechanical Design Technical Committee, the ASME
Post-Construction Subcommittee on Repairs and Testing,
and the Joint API-ASME Task Group on Fitness-for-Service.
Mr. Antaki has extensive field experience in design.
inspection, integrity, and retrofit of vessels and piping
systems and equipment. He resides in Aiken, South Carolina.

ANTAKI - Fitness For Service and Integrity of Piping-Vessels and Tanks PDF

Uploaded by

Copyright:

Available Formats

ANTAKI - Fitness For Service and Integrity of Piping-Vessels and Tanks PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ANTAKI - Fitness For Service and Integrity of Piping-Vessels and Tanks PDF

Uploaded by

Copyright:

Available Formats

A COMPLETE GUIDE TO ASME RECOMMENDED PRACTICES

FOR ASSESSING, MONITORING, ANO MAINTAINING

Fitness-for-Service and lntegrity of Piping, Vessels, and Tanks provides

Antaki V ASME and API rules for structural integrity

Visit McGraw-Hill Foucation at

Copyright © 2005 by The McGraw-Hill Companies, Inc. All rights reserved.

2.12 Types of Heat Treatment 41 3.31 Fatigue 97

5.11 Slze Effect 163 6.5 How? 214

Chapter 12. Failure Analysis 377 Preface

Appendix. WRC Bulletins 425

1.1 What ls Fitness-for-Service?

1.2 FFS and Conduct of Operations

• Is coating, cathodic protection, or lining required?

2.2. Detailed Design • Is preoperational testing and turnover to operations satis-

• Is fitness-for-service based on competent engineering, or on]y

1.3 Fitness-for-Service of Old

based on fitness-for-service of a component with the defect; instead,

Figure 1.6 Burst test with

1.5 FFS in Construction Codes

address the leak tightness of mechanical joints (threaded fittings,

1.1 O The Five Disciplines

• SQNFoU A Procedure for Safety Assessment of Components with

1.13 Cum Laude I 1 I

2. Fitness-for-Service Engineering Assessment Procedure I 1 I

5. Assessment of Local Metal Loss I 1 I

9. Assessment of Cracklike Flaws r -- I 1

E. Residual Stresses in a Fitness-for-Service Evaluation 1.15 Response Ti me

1.1 4 Technical Basis

2.1 Oemand and Capacity

2.2 Material Groups

• Cast irons: white, gray, ductile.

TABLE 2.1 Common ASTM or ASME 11 Boller Materials67

Material Tube Pipe Casting Forging Piate

Alloy steel is steel, other than stainless steel, with a mtmmum

TABLE 2.2 Common ASTM Steel Groups

Material Pipe or tube Casting Forging Plate

Nickel alloys. Nickel is a metal with good corrosion resistance, and

brazed, or soldered. lt is generally resistant to corrosion in potable

2.6 Basis for Material Selections

• ~tren~h is ~he material's yield stress Sy (elastic-to-plastic transi-

2.9 Phase Diagram of Carbon Steel

Figure 2.8 Body-cent.ered cubic

Figure 2.11Simplified view of

condition. The microstructure, depicted in Fig. 2.11, will comprise the

• Zone 1 is sufficiently rcmoved from the weld; its microstructure

2.1 O Heat Treatment

• Preheating the metal before welding, causes the weld, surrounded

Figure 2.12 Heating blankets for preheat and interpass

Maximum hardness <Brinell> Preweld beat, °F Postweld heat, °F

6 8 13 • Attach thermocouples to the component to monitor temperature ver-

2.13 Shop and Field HeatTreat ment

Figure 2.16 Field heat treatment of vessel with burners <TC:

Figure 2.18 Postweld heat trcatment.68

2.14 The larson-Miller Parameter

where T = temperature, °F, and t = time, h.

2.15 Heat and Lot

2.16 The Three Strength Parameters

where cr = engineering stress, psi

where k = strength coefficient, psi, and n = strain hardening coefficient.